UN

 SUPERFUND '87
FUN
F
RFUND '87 SUPER
PERFUND '87 SU
SUPERFUND '87 S
17 SUPERFUND

-------
      SUPERFUND  '87
              Proceedings of
      The 8th National Conference
                   (formerly)

    MANAGEMENT OF UNCONTROLLED
         HAZARDOUS WASTE SITES
     NOVEMBER 16-18, 1987 • WASHINGTON, D.C.
                   Sponsored by

The Hazardous Materials Control Research Institute
                    AFFILIATES

              U.S. Environmental Protection Agency
                U.S. Army Corps of Engineers
                  U.S. Geological Survey
           U.S. Agency for Toxic Substances & Disease Registry
               American Society of Civil Engineers
              Association of Engineering Geologists
                U.S. Department of Defense
             National Environmental Health Association
                 National Lime Association
            National Solid Waste Management Association

-------
                                           PREFACE
  The rigorous control and management of hazardous
materials  and wastes is urgently necessary for safe-
guarding  the public health,  our  environment  and
natural resources, while at the same time, fostering the
continued economic growth of the nation. Since 1981,
the Hazardous Materials Control Research Institute has
organized an annual  conference and  exhibition  to
review, update, and exchange information on the latest
research and technical findings from the laboratory, in-
dustry, and the field. The impact of federal and state
policies as well as the legal and economic issues  of
hazardous waste management are also covered in this
conference. With the cooperation of our affiliates, the
annual  Superfund  Conference  and  Exhibition  has
become the most comprehensive gathering and informa-
tion exchange forum available. This year, the title of the
Proceedings was changed from The Management of Un-
controlled Hazardous Wastes to SUPERFUND, a term
used in the authorizing legislation of 1986, and increas-
ingly recognized as the umbrella term for the complex
issues surrounding hazardous waste management.
  CERCLA  (The Comprehensive Environmental Re-
sponse Compensation and Liability Act) or  "Super-
fund," as it is now commonly known, was first passed
in 1980. This  Trust Fund, administered by the U.S. En-
vironmental Protection Agency (EPA),  was created to
help  pay  for clean up of hazardous waste sites that
threatened the public health or environment. In 1980,
EPA  had inventoried approximately 16,000 potential
sites. To be eligible for cleanup under "Superfund," a
site had to be on the  National Priorities List (NPL).
  In October 1981, EPA published an interim National
Priorities List of 115 sites  which grew to 419 in 1983.
Each year, new sites  were  proposed  and the NPL was
updated. By 1987, there were more than 1000  sites in-
cluded or proposed  for the National Priorities List.
Remedial investigations and feasibility studies  (RI/FS)
have been conducted at more than 300 NPL sites. Sixty
sites are being cleaned up and emergency cleanups have
been initiated at about 400 NPL and non-NPL sites.
There are some NPL sites that are being cleaned up by
private groups under EPA supervision.
  Under CERCLA, EPA developed a strategy composed
of three major elements. The first called for assessing
the uncontrolled hazardous waste sites in the Agency's
current inventory. Second, those sites which presented
an imminent threat to public health or the environment
were to be  stabilized. Third, using  the National Con-
tingency Plan for guidance, the NPL sites were ranked
to receive priority attention for remedial cleanup action.
  The "Superfund" extension,  the Superfund Amend-
ments and  Reauthorization Act (SARA), signed into
law in October 1986, was  funded at  a level of 9 billion
dollars.  The extension represented  a much increased
funding  level  over the  previous  five-year period,
1980-1985. A significant portion of these resources will
be devoted to remedial construction  projects at existing
and newly listed NPL sites.
  SARA is  designed to achieve greater effectiveness by
intensifying all activities under CERCLA and adding
more facets to the scope of Superfund activities. Within
the total program, SARA states will be placed in the im-
plementing  role  and greater  responsibilities will  be
delegated  to  the  EPA  Regional  Administrators.
Through the implementation of SARA, new sites will be
identified and new  technologies will be developed and
employed.
  This year's  Proceedings  contains  more than 115
papers and seminar outlines that emphasize the latest
developments and cumulative experiences gained from
the spectrum of Superfund activities. This knowledge
and experience can serve  as an immediate technology
transfer applicable to your areas of  concern.
                                                                                            Hal Bernard
                                                                                       Executive Director
                                                                                                HMCRI

-------
                           ACKNOWLEDGEMENT
  HMCRI would like to express appreciation to all of the individuals and organizations who assisted in the de-
velopment of the SUPERFUND '87 program, the Proceedings and the success of the 8th National Conference
and Exhibition on the Management of Uncontrolled Hazardous Waste Sites.
  Affiliated organizations include:
    U.S. Environmental Protection Agency
    U.S. Army Corps of Engineers
    U.S. Geological Survey
    U.S. Department of Defense
    U.S. Agency for Toxic Substances and Disease Registry
    American Society of Civil Engineers
    Association of Engineering Geologists
    National Environmental Health Association
    National Lime Association
    National Solid Waste Management Association
  The professionals on the Program  Review  Committee reviewed an unprecedented number of abstracts to
develop this informative and interesting program. The Committee was composed of:
    Hal Bernard, Hazardous Materials Control Research Institute
    Ed  Earth, U.S. EPA, Land Pollution Control Division
    John Cunningham, U.S.  EPA, Hazardous Site Control Division
    Andre  DuPont, National Lime Association
    Carolyn Esposito, U.S. EPA, Releases Control Branch, HWERL
    Patricia L.D. Janssen, U.S. Department of Defense
    Terry Johnson, National Environmental Health Association
    Don Kraft, U.S. EPA, Emergency Response Division
    Paul Lancer, U.S. Army Corps of Engineers
    Walter Leis, Association of Engineering Geologists
    Robert Olexsey, U.S. EPA, Alternative Technologies Division
    Suellen Pirages, National Solid Waste Management Association
    Stephen Ragone, U.S. Geological Survey
    Jerry Steinberg, Hazardous Materials Control Research Institute/Water and Air Resources
    Andres Talts, American Society of Civil Engineers/Defense Environmental Leadership Project
    Robert Williams, Agency for Toxic Substances and Disease Registry
  Special thanks should go to Dr. Gary Bennett, Professor of Biochemical Engineering at the University of
Toledo,  who worked as the Technical Editor, and Judy Bennett, the Editorial Consultant. Special thanks should
also be given to members of the staff of HMCRI who worked on the Proceedings and to the team of typesetters,
designers, and proofreaders who worked hard to produce this  publication. This edition of the Proceedings was
coordinated by HMCRI's Executive Director,  Hal Bernard; Janet Terner, Director of Publications; and Joyce
Wilsie, Conference Manager for SUPERFUND '87.
                                                                                                     m

-------
GLOSSARY OF FREQUENTLY USED ACRONYMS & ABBREVIATIONS
  ACS          American Chemical Society
  AICE         American Institute of Chemical Engineers
  API          American Petroleum Institute
  ATSDR       U.S. Agency for Toxic Substances and Disease Registry (Atlanta, GA)
  CAA          Clean Air Act
  CERCLA      Comprehensive Environmental Response Compensation and Liability Act
  CMA         Chemical Manufacturers Association
  COE          U.S. Army Corps of Engineers
  CWA         Clean Water Act
  DOE          U.S. Department of Energy
  DOI          U.S. Department of the Interior
  DOT          U.S. Department of Transportation
  EDF          Environmental Defense Fund
  EMSL         U.S. EPA Environmental Monitoring Systems Laboratory
  EPA          U.S. Environmental Protection Agency
  FEMA        U.S. Federal Emergency Management Agency
  HMCRI       Hazardous Materials Control Research Institute
  HMTA        Hazardous Materials Transportation Act
  HRS          Hazardous Ranking System
  HSWA        Hazardous and Solid Waste Amendments
  HWERL       U.S. EPA Hazardous Waste Engineering Research Laboratory
  HWTC        Hazardous Waste Treatment Council
  LUST         Leaking Underground Storage Tanks
  MCL          Maximum Contamination Level
  NIOSH        National Institute for Occupational Safety and Health
  NOAA        National Oceanographic and Atmospheric Administration
  NPL          National Priorities List
  NRDC        National Resources Defense Council
  NSWMA       National Solid Waste Management Association
  NWA         National Water Alliance
  OERR         U.S. EPA Office of Emergency and Remedial Response
  ORD          U.S. EPA Office of Research and Development
  OSW          U.S. EPA Office of Solid Waste
  OSWER       U.S. EPA Office of Solid Waste and Emergency Response
  OTA          U.S. EPA Office of Technology Assessment
  PCB          Polychlorinated biphenyl
  PRP          Potentially Responsible Party
  RCRA         Resource Conservation and Recovery Act
  RI/FS         Remediation Investigation/Feasibility Study
  ROD          Record of Decision
  SARA         Superfund Amendments and Reauthorization Act of 1986
  SITE          Superfund Innovative Technology Evaluation program
  TSCA         Toxic Substances Control Act
  TSDF         Treatment, Storage, Disposal Facility
  TTU          Transportable Treatment Unit
  UST          Underground Storage Tank
  USWAG       Utility Solid Waste Activities Group

-------
                                                  CONTENTS
Preface	

Acknowledgements

Glossary	
.in

 .v
      EPA POLICY PAPERS AND GUIDELINES
Initiatives for Improving RI/FS Project Performance	1
Lisa G. Feldt, J. Steven Paquette and Wendy
  L. Sydow
The Potentially Responsible Party Search 	5
Donna Gerst, John Ely, Esq. and William Heglund
Mass Notification of Owners Under the Defense
Environmental Restoration Program	7
H. David Hill, Timothy E. Kelleher and Gerald J.
  Stadler, P.E.
Status of the U.S. EPA's Pre-Remedial Program	14
James R. Jowett, Scott C. Fredericks and Lucy Sibold
Meeting the New OSHA Waste Training Requirements	18
Martin G. Kemplin and Philip L. Jones
Potentially Responsible Party Search Methodologies	21
Donna Lee Gerst and Laurie A. Redeker
Superfund Innovative Technology Evaluation
(SITE) After the First Year	25
Ronald D. Hill, P.E.
A Compendium of Superfund Field Operations
Methods	28
James B. Moore, Robert E. Stecik, Jr. and Lisa Feldt
                      LIABILITY

Federal and State Liability Standards for
Superfund Response Actions Contractors	34
Robert J. Mason, Douglas W. Kohn, J.D. and Mark
  F. Johnson
Calculating a Risk Premium for a CERCLA Site	41
Yardena Mansoor and Thomas Gillis
Historical Risk Assessment of Environmental
Liabilities at Former Industrial Properties	45
Sandy Peterson
Corporate Successor Liability for Environmental Torts	48
Mell J.-Branch Roy, Esq.
                POLICY ASSESSMENT
State and Local Jurisdiction at Federal Facilities
with Hazardous Waste Sites	53
John E. Cromwell, Ph.D. and Alyson A. Hennelly
Risk at Superfund Sites: A National Perspective	56
Craig Zamuda, Ph.D.
Risk Assessment in Superfund: Policies
and Procedures	61
Sherry Sterling and Craig Zamuda, Ph.D.
                    SAMPLING AND MONITORING
                                                              .63
                                                              .66
                                                              .72
Selecting a Contract Laboratory	
Craig W. Rice
The Use of Short-Term Bioassays to Assess
Clean-Up Operations of Sites Contaminated with
Hazardous Wastes	
Phebe Davol, Kirby C. Donnelly and Kirk W. Brown
Application of Data Quality Objectives at a
Superfund Remedial Investigation at a Former
Municipal Landfill	
Marc P. Lieber, John D. Frost, Paula M. Lia
  and Michael Amdurer, Ph.D.
Remediation at a Major Superfund Site—Western
Processing, Kent, Washington	
Kenneth A. Lepic, P.E. and Allan R. Foster
Low Level Volatile Organic Analysis Utilizing a
Closed Loop Stripping Methodology as Compared to
U.S. EPA IFB CLP Analysis Utilizing Purge
and Trap Methodology at a CERCLA Site	
Michael S. Zachowski
State Monitoring Well Regulation: Need for Consistency	89
Charles A. Job and Gilbert Gabanski
                                                              .78
                                                                                                                   .85
          Quality Assurance for the Field Laboratory,
          B. Chris Weathington
                                                     .93
                           FIELD SCREENING
          Effects of Environmental Variables on Soil
          Gas Surveys	
          Louis S. Karably, P.E. and Kevin B. Babcock
          Assessing the Validity of Field Screening of Soil
          Samples for Preliminary Determination of
          Hydrocarbon Contamination	
          Philip G. Smith and Stephen L. Jensen
                                                     .97
                                                    .101

-------
 Rapid Soil Extraction and Cleanup Procedure
 for PCBs	104
 Jon C. Gabry, Ph.D.
 Field Analytical Screening for Acid Extractables
 In SoU and Water	107
 Roger N.  McGinnis, Ph.D. and Andrew J. Hafferty
       MODELING OF HAZ MAT TRANSPORT
 Computer Modeling Results and Groundwater
 Treatment System for Price's Landfill No. 1 	
 Abu M.Z. Alam, Sc.D., P.E., George Klein,
   Salvatore Badalamenti and Robert McKnight
 Trend-Surface Modeling of Groundwater Data	
 Charles T. Kufs
 The Impacts of Using Assumed Versus Site-Specific
 Values in Determining Fate and Transport	
 Pinaki Banerjee, Ph.D. and David H.  Homer, Ph.D.
 Effects of Diffusion Upon Multi-Dimensional
 Contaminant Transport	,
 Maj. Mark N. Goltz (USAF), Ph.D., P.E. and
   Paul V. Roberts, Ph.D.
               HEALTH ASSESSMENT
 Regulation of Dioxin as a Carcinogen: A National
 and International Dilemma	
 Norbert P. Page, D.V.M.
 Exposure and Public Health Risk Assessment for
 the Balrd & McGnire Snperfund Site	
 David E. Burmaster, Ph.D., Scott K. Wolff, John
  J. Gushue, Esq., Brian L. Murphy, Ph.D. and
  Charles  A. Menzie, Ph.D.
 The Human Health Risks of Recreational Exposure
 to Surface Waters  Near NPL Sites: A Scoping
 Level Assessment	
 Gary K. Whitmyre, James J. Konz, Mark L. Mercer,
  H. Lee Schultz and Steve Caldwell
 Environmental Modeling and the Snperfund
 Exposure Assessment Process	,
 Peter Tong, H. Lee Schultz and Seong Hwang, Ph.D.
An  Exposure Assessment ModeUng System  for
Hazardous Waste Sites 	
Alison C. Taylor, David E. Burmaster, Ph.D., Brian
  L. Murphy, Ph.D. and Scott H. Boutwell
Human Exposure Potential Ranking Model	
Lee Ann Smith,  Cynthia D. Patrick and Charles
  M. Hudson
Hazardous Waste Site Health and Safety after OSHA  .
Martin S. Mathamel
Assessing Risk from Dermal Exposure at Hazardous
Waste Sites 	
Elizabeth A. Ryan, Elizabeth T. Hawkins,  Brian
  Magee, Ph.D. and Susan L. Santos
.111
.120
.126
.129
.132
.139
.143
.149
.153
.15*
.162
.166
Chemical Oxidation Destruction of Organic
Contaminants In Groundwater	I7*
Donald G. Hager, Carl G. Loven and Christopher
  L. Giggy
MobUe Waste Oil Recovery	179
Craig A. Nowell and Mark J. Hardy
Electrochemical Oxidation of Hexone
and Other Organic Wastes  	1*3
Alex G. Fassbender,  Peter M. Molton, Ph.D.
  and Greg Broadbent
Advanced Chemical Fixation of Organic Content
Wastes In Conjunction with Japanese Ground
Engineering Equipment	187
Jeffrey P. Newton
Bloremedlation of Contamination by Heavy
Organics at a Wood Preserving Plant Site	193
Ronald J. Linkenheil and Thomas J. Patnode
SoU Stabilization TreatabiHty Study at the
Western Processing Superfund Site	198
John J. Barich, Joseph Greene and Rick Bond


      CONTAMINATED AQUIFER CONTROL
Groundwater Restoration at McCkUan AFB	204
Mario lerardi, Paul Brunner and Edward J.
  Cichon, Ph.D.
"Decay Theory" Biological Treatment for
Low-Level Organic Contaminated Groudwater
and Industrial Waste	208
Kevin M.  Sullivan and George J. Skladany
Evaluation of Groundwaler Remediation
Techniques for Fractured  Bedrock Using Aquifer
Response Tests	213
Kenneth A. Wallace, Paul L. Kannazinski and
  Douglas J. Yeskis
Gronndwater Treatment for Mixed Contaminants:
Processes and Facilities	218
Thomas M. Sanders, P.E.
Effective Startup and Operational Procedures
for Groundwater Remediation Systems	223
David W. Hale, Marc J. Dent and David G.
  Van Arnam, P.E.
Development of Methodologies  for Evaluation
of Well-Point Systems	228
T.Y. Richard Lo, Ph.D..  P.E. and Victor Owens
Comparison of Pollutant Fluxes In Saturated
and Unsaturated Flows Beneath Hazardous
Waste Sites  	231
Lance R.  Cooper and Frank L.  Parker, Ph.D.
Remedial  Action Evaluation System: A Methodology
for Determining the Completion Point for
Aquifer Restoration Programs	238
Paris Hajali, Ph.D.
                    TREATMENT
On-Slte Water Treatment Under Superfund:
Economical and Effective	
Jeffrey S. Clark and Peter D. Neithercut
Optimization of Remedial Treatment Actions
for Contaminated Soils	
Edwin F. Barth, P.E.  and John J. Barich
.169
.172
                           RISK ASSESSMENT
Using Risk Concepts In Superfund	
Joel S.  Hirschhorn, Ph.D., Kirsten U.
  Oldenburg and David Dorau
                                                              .251

-------
Risk Communication: A Critical Part of the
Public Participation Process	254
Susan L. Santos and Sally Edwards
Procedure to Assist Decision-Makers in Selecting
a Remedial Alternative for Hazardous Waste Sites	258
Ram Swaroop, Ph.D. and Richard Carter, P.E.
International Study of the Social, Psychological
and Economic Aspects of Problem Hazardous
Waste Sites	264
Michael A. Smith and Sara B. Baker
                                                    .268
                                                    .273
                                                    .280
                                                    .284
               VOLATILE ORGANICS
           MONITORING REMEDIATION
Management of Waste Compressed Gas Cylinders ..
Charles Mattern and Dan Nickens
Vacuum Extraction of Hydrocarbons from
Subsurface Soils at a Gasoline Contamination Site ..
Joseph Applegate, John K. Gentry, P.E.
  and James J. Malot,  P.E.
Analysis Methods and Quality Assurance
Documentation of Certain Volatile Organic
Compounds at Lower Detection Limits 	
Susan Sherman, Ward Dickens and Harold Cole
Toxic Air Quality Investigation at a
Hazardous Waste Site	
George D. Marquardt
                 SITE REMEDIATION
Lessons Learned in Remedial Design and
Cleanup of a Federal Superfund Site	296
Thomas F. Maher, P.E., Michael W. McLaughlin,
  J.D. and Christine Beling
Role of Surface Geophysics in Developing
Buried Waste Removal Specifications	300
Norm N. Hatch, Jr., William Owens, Samuel
  Shannon and Patricia Markey
Immediate Removal Activities at a Dioxin-
Contaminated Mobile Home Park	306
Russell B. Krohn and James R. MacDonald
Groundwater Cleanup at Selected Superfund Sites	311
Lisa Haiges and Robert Knox, Ph.D.
Pilot-Scale Bioremediation at the Brio
Refining Superfund Site	315
Bruce S. Yare, Derek Ross, Ph.D. and David
  W. Ashcom
Evaluation of Well Field Contamination Using
Downhole Geophysical Logs and Depth-Specific
Samples	320
George T. Ring and Thomas C. Sale
A Phased  Approach to Remedial Investigations:
Focusing Effort and Reducing Overall Remedial
Investigation Costs	326
Gary Hoffmaster, Leonard C. Johnson, Jane M.
  Patarcity and Robert J. Hubbard
Verona Well Field—Where the Action Is	330
John C. Tanaka, Joan van Munster, P.E.
  and Alan Amoth, P.E.
Impervious Liner Installation Along Canal Bottom	334
Eugene F. Stecher, P.E.
Analysis of RCRA Closure Options for
Superfund Site	337
John M. Cunningham and Marlene G. Berg
An Innovative Approach for Remediation of
Metal-Contaminated and Environmentally
Sensitive Marsh Areas	341
Hsin H. Yeh, Ph.D., P.E., Dev R. Sachdev,
  Ph.D., P.E. and Joel A. Singerman
Hydrogeologic Assessment, Delineation and
Remediation of a Shallow Groundwater
Contaminated Zone	348
Martin M. Fontenot
The South Valley San  Jose 6 Superfund Site:
Special Issues and Problems	355
Kathleen O'Reilly, Steve Tarlton and Paul Karas
Installation of Monitoring Wells Using the Dual
Wall Hammer Drilling Technique	358
Thomas C. Sale and Sara E. Rhoades
Innovative, Fast-Track Multi-Media PCB Cleanup	362
Kevin Chisholm, P.E.
Design of Remedial Measures and Waste Removal
Program, Lackawanna Refuse Superfund Site	367
Marcella J. Blasko, Beth F. Cockcroft, William C.
  Smith and Patrick F. O'Hara
Horizontal Radials for Geophysics and
Hazardous Waste Remediation	371
Wade Dickinson, R. Wayne Dickinson, Peter A.
  Mote and Jerome S. Nelson
A Model for Estimating the Cost of Superfund
Remedial Actions	376
Richard K. Biggs and  R. Benson Fergus, P.E.
                                                                      CONTAMINATED SOIL TREATMENT
                                                                                                                 .380
                                                                                                                 .385
                                                                                                                  .390
                                                                                                                  .396
Heavy Metals-Contaminated Soils Treatment	
Peter S. Puglionesi, P.E., Jaisimha Kesari,
  Michael H. Corbin, P.E. and Erik B. Hangeland
A New Method to Characterize Contaminated Soils.,
Namunu J. Meegoda, Ph.D. and Prasanna
  Ratnaweera
In Situ, Vacuum-Assisted, Steam Stripping  of
Contaminants from Soil	
Arthur E. Lord, Jr., Ph.D., Robert M. Koerner,
  Ph.D., P.E., Vincent P. Murphy and John E.
  Brugger, Ph.D.
Innovative In Situ Decontamination System  	
Richard Van Tassel, Ph.D. and Phillip N. LaMori
                                                                           RAD AND MIXED WASTES
                                                              Mixed Waste and ALARA	403
                                                              Nancy P. Kirner
                                                              Overview of the West Valley Demonstration Project  	405
                                                              M.D. Weingart, R.R. Borisch and D.R. Leap
                                                              Application of the Remedial Action Priority
                                                              System to Hazardous Waste Sites on the
                                                              National Priorities List	499
                                                              Gene Whelan, Robert D. Brockhaus, Dennis L.
                                                                Strenge,  James G. Droppo, Jr., Marcia B,
                                                                Walter and John W. Buck

-------
            WATERWAYS AND WETLANDS
                    RECLAMATION
  Role of Sedimentary Channel Deposits In
  Contaminant Migration and Remediation Design	
  Marilyn A. Plitnik and Ramesh J. Shah
  Application of a Mixed-Method Analytical
  Scheme for Analysis of PCB In Water and
  Sediment Samples from a Polluted Estuary	
  Richard A. McGrath, William G. Steinhauer
    and Siegfried L. Stockinger
  Sampling Program for the New Bedford Harbor
  Superfund  Site: Modeling the Movement of
  Pen-Contaminated Sediments	
  Richard A. McGrath, Michael S. Connor, Ph.D.
    and William G. Steinhauer
  A Wetland Assessment Procedure for
  Superfund Sites	
  Anne Sergeant


                   MINING WASTES

 Degree of Cleanup: SARA  121(d), ARARs
 and Mining Sites	
 Don C. Porter and William E. Cobb
 Clearing and Abandonment of Deep Water Wells
 at the Tar Creek Superfnnd Site	
 Michael Howar, Douglas Hayes and Bart Gaskill
 Mobilization and Transport of Metals from Mine
 Tailings to an Alluvial Aquifer	
 William B. Mills, P.E., Steven A. Gherini, P.E.
  and Gary  Bigham
 Technical Approaches on the UMTRA Project
 Relevant to Superfnnd Projects	
 Jack A. Caldwell and Duane Truitt

                   INCINERATION
 Utilization of Mobile Incineration at the
 Beardatown Lauder Salvage Yard Site	
 James A. Janssen, P.E., Robert Munger, John W.
  Noland, P.E., Nancy p. McDevitt and Luis
  A. Velazquez, P.E.
Automated Analysis of Adsorbent Traps Used In
the Volatile Organic Sampling Train (VOST)	
Paul E.  Kester and Alan D.  Zaffire
Use of Mobile Incineration to Remediate
the Lenz Oil Site	
James F. Frank, Mary E. Dinkel and Desi M. Chari
Development of a Gaussian Puff Model for
Over-Ocean Incineration Applications	
James G. Droppo, Jr., Ph.D.. Richard M. Ecker
  and David  Red ford

                    MULTI-MEDIA
Risk Analysis of Pollutants at Hazardous Waste
Sites: Integration Across Media Is the Key	
Lyse D.  Helsing, Ph.D., Mary P. Morningstar, Esq.,
  Joan B. Berkowitz, Ph.D. and Thomas
  T. Shen, Ph.D.
Factors and Phenomena Involved In Multimedia
Exposure Assessment	
T. Edward Fenstermacher, Ph.D. and Luca Ottinetti
 .414
 .420
 .426
 .431
 .436
 .439
.449
.453
.457
.459
.465
.471
.476
 The Use of Stabilized Aqueous Foams to Suppress
 Hazardous Vapors—A Study of Factors
 Influencing Performance	
 Roger R. Aim, Chris P. Hanauska, Kathleen A.
   Olson and Myron T. Pike, Ph.D.
 Multimedia Approach to Risk Assessment for
 Contaminated Sediments In a Marine Environment .
 Seong T. Hwang, Ph.D.
               COST AND ECONOMICS
 Earned Value as an Appropriate Tool for
 Controlling Remedial Planning Contracts	
 James B. Chaffee. Jr., Robert T. Fellman and
   John J. Smith
 Bidding and Awarding of Remedial Construction
 Contracts for Hazardous Waste Sites	
 Patrick F  O'Hara, Beth F. Cockcroft and
   William C. Smith
 Environmental Risk Considerations In Real Estate
 Transfers for Active Waste Management Facilities ..
 Mark P. Zatezalo and Patrick F. O'Hara

                DISPOSAL/STORAGE
 A Demonstrable Performance Approach for
 Waste Isolation In Geologic Repositories	
 Daniel C. Melchior, Ph.D., Sarah Hokanson
  and Jon Greenburg
 A Numerical Evaluation System for Comparison
 of Potential Land Disposal Sites 	
 Owen S. Ruta and Geoffrey W. Watkin
 Proposed Short-Term  Burial/Storage Method
 for Unassayed Hazardous Waste	
 Arthur G. Clem, P.E.
 Impact of the RCRA Land Disposal Restrictions
 on Superfund Response Actions	
 James Antizzo, John Cunningham, Kathleen
  Hutson and Amelia Heffeman
                     SEMINARS
Natural Resource Damages Under Superfund	
Richard W. Dunford.  Ph.D. and William H.
  Desvousges, Ph.D.
Contractor Liability and Indemnification
Under Superfund	
J. Kent Holland Jr., Esq. and Robert J.  Mason
Soli-Gas Surveying for Subsurface Organic
Contamination: Active and Passive Techniques ....
H.B. Kerfoot
Introduction to Dispersion Modeling of
Hazardous Releases	
Ashok Kumar, Ph.D.
Clinical Medical Surveillance and Epidemiology:
Methods for Assessing and Protecting the Health
of Hazardous Waste Workers	
Bertram W. Carnow, M.D. and Shirley A.
  Conibear, M.D., M.P.H.
Bloremedlation of Hazardous Waste Sites	
Raymond Loehr, Ph.D., John Ryan and Ronald
  Linkenheil
                                                               .4*0
                                                               .495
                                                               .492
                                                               .496
                                                               .499
                                                               .502
                                                               .SOS
                                                               .512
                                                               .515
                                                              .517
                                                              .520
                                                               .513
                                                               .525
                                                               .532
                                                               .533
           Exhibitors
Author Index

Subject Index
.537

.551

.557

-------
                                      Initiatives  for  Improving
                                    RI/FS  Project Performance

                                                    Lisa G. Feldt
                                     U.S. Environmental Protection Agency
                                         Hazardous  Site Control Division
                                                 Washington, D.C.
                                                 J.  Steven Paquette
                                                  Wendy L. Sydow
                                      CDM Federal Programs Corporation
                                                  Fairfax,  Virginia
ABSTRACT
  The U.S. EPA has implemented several initiatives aimed at
improving performance on Superfund RI/FS projects. The im-
portance of these initiatives has been reinforced by the demanding
schedules  set for remedial activities  by SARA.  This paper
summarizes the results of a study conducted by the U.S. EPA on
RI/FS improvements. The following initiatives are discussed:

• Phased RI/FS execution
• Quality control and technical advisory committee
• Streamlined project planning
• Management of handoffs and critical activities

INTRODUCTION
  Early in the Superfund program, the U.S. EPA anticipated that
an average RI/FS would be completed in 18 months. Project com-
plexity, the program learning curve and inefficiencies in the process
have resulted in project delays and cost increases. Currently, project
planning activities take approximately 6 months to complete and
a full RI/FS runs an average of 25 months. The U.S. EPA's goal,
through implementing a series of RI/FS improvement initiatives,
is to improve the schedule and cost efficiency of the RI/FS process
while concurrently improving  the technical quality of the RI/FS
work.
  The U.S. EPA has analyzed the RI/FS process by evaluating
the critical tasks that occur during the planning, data collection
and evaluation, and engineering analysis phases of an RI/FS. Then
they modeled  various  implementation  alternatives  aimed at
improving project  efficiency, cost-effectiveness and technical
quality '.
  Modeling was conducted using a critical path method (CPM)
computer-based model. CPM is a project management tool which
frequently is used to plan schedule and resource allocations for
engineering projects. The CPM method is based on linking tasks
over a period of time where, in any given period, the lengthiest
critical task (i.e., must be completed before the project can proceed
to a subsequent task) is listed on the  "critical path." In this manner,
project managers can identify  which tasks are most important to
complete  on schedule to avoid lengthening  the overall project
schedule.
  The focus of the analysis was to  develop a realistic strategy for
RI/FS improvements that would meet the goals of reduced schedule
and costs and improved technical quality without requiring major
changes to the remedial program. Through the CPM analysis, the
U.S. EPA was able to predict significant time and cost savings
through planning a hypothetical project incorporating the improve-
                                                                                  RI/FS CPM SCHEDULE COMPARISON

Biullnt CPU

Modified
Uontht

Planning HI Ft I POST-F* 1

Plinnlng Rt Phซซ. 1 Rl Ph.t* || FS PO1T-FS 1

                  RI/FS COST/RESOURCE COMPARISON

BMtlllM CPU



Coปl (JI.OOO'l)

P,.n,,n, .1 F. 'ซ'•



50 100 150 200 250 300 350 400 450 500 550 600 650 700 750 BOO
                         Figure 1
             CPM Schedule and Cost Comparison

ment initiatives described in this paper. The complete original and
modified CPM networks are too large to be displayed in this
document.
  Figure 1 illustrates the schedule and cost reductions realized
through incorporation of program improvements into the CPM
network. The figure shows the remedial investigation split into two
discrete data collection activities,  while still reducing the overall
schedule and costs. The feasibility study may appear unrealistically
short unless one realizes that it is actually initiated and conducted
concurrently with the remedial investigation during development
and evaluation of alternatives. The post  feasibility study phase
includes activities from predesign through project closeout (work
assignment completion report).
  The Superfund improvement objectives which the  U. S. EPA
hopes to achieve through implementation of these improvement
initiatives include: (1) containing project planning activities within
a 3-month period after project initiation,  (2) ultimately reducing
the overall RI/FS process to an 18-month schedule, (3) reducing
overall costs, and (4) improving technical quality of RI/FS projects
2. The sections below describe the U. S. EPA's evaluation of the
current RI/FS process and the program improvement areas: phased
RI/FS, quality control, streamlined project planning and manage-
ment of handoffs and critical activities.

EVALUATION OF THE CURRENT RI/FS  PROCESS
  RI/FS execution is affected by many elements that have little
to do with the technical analysis needed to select a remedial alter-
                                                                           EPA POLICY PAPERS AND GUIDELINES    1

-------
 native for a site. Schedule and cost increases sometimes are caused
 by delayed deliveries of validated data, lengthy review times or col-
 lection of insufficient or inappropriate data during the RI. These
 elements and others were evaluated by collecting data on completed
 RI/FS projects and analyzing the impacts of various delays on to-
 tal project duration. In addition to using data from completed fund-
 led projects, the analysis also relied on input from over 50 individu-
 als in seven U. S. EPA Regions. The review of the existing RI/FS
 process also included an analysis of handoffs (transfers of respon-
 sibility) and  their contribution to project delays. Regional input
 was gathered using a questionnaire which focused on identifying
 chronic project delay areas and successful program improvement
 efforts.
   The analysis of the existing RI/FS process indicated that delays
 in the process are largely attributable to the following factors:

 • Administrative and document review handoffs (approximately
   20 on a typical RI/FS)
 • Sample analysis  and  data validation
 • Redundancy in project planning documents
 • Lack of standardized formats for planning documents
 • PRP negotiations

   Good communication between contractors and the U. S. EPA
 was identified as a consistent  factor in successful projects. In
 particular, project  review meetings were felt to give contractors
 direction, keep  projects on track, identify problems early and
 generally result  in improved performance.
   Based on the analysis, the U. S. EPA  developed recommenda-
 tions for program improvement in the following areas:

 • Phased RI/FS execution
 • Quality control and technical advisory committee streamlined
   project planning
   •  Management of handoffs and critical activities

   Recommended activities are listed  for each program improve-
 ment area in  the sections below. Some activities are listed under
 more than one improvement area. For example, early involvement
 of a Technical Advisory Committee  can be instrumental in im-
 proving the phased RI/FS, quality control and streamlined project
 planning. Though the program improvement initiatives fall into
 the general areas listed above, they really cannot be separated from
 one another and  should be viewed as an integrated package.

 PHASED RI/FS EXECUTION
   The RI/FS  is evolving into a more interactive process which leads
 toward phasing the RI and FS (3). In a phased RI/FS, the results
 of each phase are evaluated and used to  define the more focused
 scope  of subsequent phases,  thereby  minimizing extraneous
 activities. This leads to more efficient data collection and evalua-
 tion  efforts.  Each phased RI/FS process should consist of the
 following steps:

 •  Scoping of the RI/FS - Collect and analyze existing data, iden-
   tify potentially applicable or relevant and appropriate require-
   ments (ARARs), determine initial action levels, identify initial
   data quality objectives (DQOs) for data collection activities and
   identify operable units or response scenarios.
 •  Site Characterization - Conduct field  investigation, define the
   nature and extent of contamination, conduct baseline  risk
   assessment and refine remedial action goals.
 •  Development of  Alternatives   Identify potential treatment
   technologies and containment/disposal requirements, screen
   technologies, identify action-specific ARARs and develop a range
   of alternatives.
 •  Initial Screening of Alternatives - Screen alternatives to identify
   those that  should be evaluated in detail.
 •  Treatability Studies - Perform bench or pilot tests as necessary.
 •  Detailed Analysis of Alternatives - Develop performance criteria;
   analyze relative costs, long- and short-term effectiveness and

2    EPA POLICY PAPERS AND  GUIDELINES
   implementability; and verify/compare protection of public health
   and the environment, compliance with ARARs, reduction of
   mobility and.or toxicity and other statutory factors.

   Proper planning is important to successful execution of a phased
RI/FS. Successful planning requires closely coordinating develop-
ment of the RI with the FS to assess data needs for evaluation of
alternatives and any necessary treatability studies.
  Development of data quality objectives (DQOs) is an integral
part of the planning process for each phase of data collection
activity 4'5.  DQOs are qualitative and  quantitative statements
which specify the quality of the data required for specific use*.
DQOs are established during project scoping and at the initiation
of any subsequent data collection activities to ensure that data are
of sufficient quality for their intended use. DQO development is
integrated with the project planning process, and the results are
incorporated into the sampling and analysis (S&A) plan, quality
assurance project plan (QAPP) and, in general terms, the work
plan. The DQO  process results in a well thought out S&A plan
which details the selected sampling and analysis approach. All field
investigation activities should be conducted and documented in a
manner that ensures that sufficient data of known quality are col-
lected to support  sound decisions concerning remedial action selec-
tion. This applies to fund-led. Federal or State enforcement-led
and potentially responsible party-lead projects.
  Phasing the RI/FS process, with incorporation of DQOs before
the initiation of each stage of data collection, ensures that field
activities are focused to obtain necessary data in the most efficient
manner.
  The following results of implementing a phased RI/FS approach
should yield project improvements:

• Field activities can be initiated earlier. Because project
  planning can focus on initial RI/FS phases, subsequent phases
  only need to be discussed in general terms in the initial work plan.
• Initial site visits and limited field sampling can be used to help
  define the work plan and site conceptual model more dearly from
  the beginning of the RI/FS process.
• Resources can be more effectively managed through using DQOs
  to guide data collection activities and working from a very
  focused scope at each phase of the work.
• Unfeasible remedial alternatives can be eliminated earlier in the
  RI/FS process through  evaluation of  results from initial data
  collection activities.
• Overall RI/FS schedule and costs can be reduced.
• Treatability  studies  can be conducted  during the RI/FS, con-
  current with other activities.

QUALITY CONTROL AND TECHNICAL
ADVISORY COMMITTEE
  The U. S. EPA wants to ensure that the RI/FS process  is con-
ducted in a manner that yields high-quality products and is com-
parable and consistent among the Regions. This goal can be
attained through implementation of Region, specific technical
quality control processes which would include involvement  of a
Technical Advisory Committee (TAC) at key project milestones.
A TAC is a group of senior level U. S. EPA/State and contractor
personnel selected to serve as technical reviewers for a project based
on their areas of expertise. Review of key deliverables in the draft
stage can improve the quality of the documents and expedite the
U. S. EPA intra-Agency reviews. The Remedial Project Manager
(RPM) is expected to  review all deliverables. Early TAC review
will also facilitate better management and more realistic cost projec-
tions for  RI/FS projects through early identification of technical
and policy issues. The following quality control and technical review
activities can be  important components of a  program of project
improvement:

• Identify and draw upon technical experts and/or other Agencies
  (e.g., Bureau of Mines  for a mining site) or within  the U. S

-------
  EPA for an early brainstorming session to review the overall
  scope of the project, identify technical or policy issues and assess
  options for reducing project costs. Information from this session
  would then be made available to the TAC for consideration in
  their review of project deliverables.
• Identify and convene a TAC at project milestones; e.g., draft
  work plan, preliminary summary of site investigation (first phase
  RI), second phase RI, initiation of FS phase and predesign.
• Clearly identify reviews and signoffs required for deliverables
  in the Region.

STREAMLINED PROJECT PLANNING
  Project planning encompasses the period from work assignment
initiation through approval of the work plan. For past RI/FS
projects, these activities have taken an average of six months and
a significant portion of the project budget to complete. A major
hurdle to getting a fully approved work plan has been reaching
agreement on major scope items scheduled a year or two in the
future. The analysis of the RI/FS process examined ways to allow
projects to: (1) be initiated more rapidly, (2) involve key reviewers
at critical  stages in the planning process, (3) reduce the number
of documents and redundancy in documents, and (4) allow work
to proceed while final work  plan  negotiations  are completed.
Efforts to streamline project planning were focused on completing
project planning activities within three months.
  The following activities are recommended for streamlining the
project planning process:

• Consolidate S&A plans and QAPPs - As stand-alone documents,
                           Table 1
          Management of Handoffs and Critical Activities

    Task                 Title

      1            Project Planning

      2            Community Relations

      3            Field Investigation

      4            Sample Analysis/Validation

      5            Data Evaluation

      6            Assessment of Risks

      7            Treatability Study/Pilot Testing

      8            Remedial  Investigation Reports

      9            Remedial Alternatives Screening

      10           Remedial Alternatives Evaluation

      11            Feasibility Study (RI/FS) Reports

      12           Post RI/FS Support

      13           Enforcement Support

      14           Miscellaneous Support

      15           ERA Planning
  these plans often are redundant and therefore require duplication
  of contractors' efforts.
• Incorporate standard procedures by reference - This will avoid
  repeating technical reviews of a procedure that has already been
  approved for use in the Region. The Compendium of Field Oper-
  ations Methods6 presents standard procedures for a variety of
  typical RI activities.
• Use standardized tasks - The Agency has developed a standard
  task structure which can be used for all RI/FS studies7. This
  task structure will standardize cost and schedule tracking and
  allow development of a data base that will be used to better
  estimate resource requirements for new remedial projects. The
  standard tasks are listed in Table 1.
• Minimize intra-Agency reviews - Rely on contractors, internal
  quality control/review procedures and RPM review for  most
  project documents and limit intra-Agency reviews  to the work
  plan, community relations  plan and S&A plan.
• Initiate  preliminary site work upon interim authorization -
  Authorization for initiation of Phase I field activities can be
  granted upon approval of the Phase I S&A plan, possibly be-
  fore approval of the complete work plan.
• Make work plans specific for initial phases of work, general for
  later phases - This will expedite development and review of the
  work plan; subsequent changes in the technical direction of the
  work can be documented through the use of a Technical Direc-
  tion Memorandum (TDM) as long as the work is within the origi-
  nal scope and budget of the  assignment. Guidance on TDMs can
  be found in the Federal-led RPM Handbook 8. The State-led
  RPM Handbook 9 also contains valuable project management
  guidance.
• Incorporate Technical Advisory Committee (TAC) review into
  project planning phase.

MANAGEMENT OF HANDOFFS
AND CRITICAL ACTIVITIES
  Critical activities have the potential to directly affect a project
schedule. That is, if certain tasks cannot be started until another
task is completed, then any delay in the first task will delay the
second task (e.g., data validation cannot begin until sample analysis
results are available). When delays occur with these activities, the
completion of the entire RI/FS inevitably is postponed. These are
exactly the types of events that this analysis has sought to identify
and resolve. It is very important,  therefore, to avoid  delays in
critical activities.
  A handoff is any transfer of responsibility for administrative
or technical project activities. Examples include turning samples
into CLP for analysis, turning data over to BSD for validation
or submitting documents to various U. S. EPA offices for intra-
Agency review. Project handoffs have been identified as a major
cause for delays in the RI/FS process, particularly for critical
activities. Increased State participation in the RI/FS process may
create additional handoffs for some projects. Care must be taken
to ensure that these reviews and other new coordination activities
required by SARA are managed carefully to avoid project delays.
  Since handoffs have been identified as a key source of project
delays, they should be kept to a minimum and,  when possible,
should be  planned so that  they do not  delay  critical project
activities. Critical activities can be minimized by scheduling con-
current activities, by receiving interim approvals,  and by phasing
critical field tasks. Recommendations for management of handoffs
and critical activities include the following:

• Turn data over to contractor for pre-analysis prior to data vali-
  dation, thereby avoiding delays in data evaluation. This proce-
  dure will enable contractors to initiate technology screening and
  development of alternatives earlier and initiate scoping for sub-
  sequent  phases.  However, unvalidated  data  should not  be
  released to other organizations, Agencies or individuals.
• Allow contractors to validate CLP data according to the U. S.
                                                                              EPA POLICY PAPERS AND GUIDELINES

-------
   EPA standard procedures with Regional audits.
 •  Minimize intra-Agency document reviews. (The RPM will held
   responsible for review of all deliverables).
 •  When a responsibility transfer is necessary, obtain a commit-
   ment from the receiving party to meet schedule requirements.
 •  Involve key decision makers in the TAG meetings to reduce
   review time of RI/FS deliverables.
 •  Initiate treatability and/or pilot testing during the RI (after initial
   alternatives evaluation) to assist in remedial alternative selection
   and expedite predesign.
 •  Use results from the analysis of screening samples to develop
   a conceptual model for the site and to perform preliminary tech-
   nology and alternative screening.
 •  Provide interim approvals for initial field activities prior to full
   work plan approvals to expedite data collection and analysis.

 FUTURE PROGRAM INITIATIVES
   Additional Superfund program initiatives are under way which
 will help further reduce project schedules and provide project
 management assistance. Among these efforts are revised award fee
 procedures and the alternate remedial contracts strategy (ARCS).
 OERR also has initiatives under way to develop a nationwide treat-
 ability/pilot study subcontracting support system and to develop
 additional technical  standard procedures.
   Revised award fee procedures give the Regions more control over
 evaluation of contractor performance I0. In addition, an initiative
 is under way to evaluate various strategies for streamlining the sub-
 contracting process.  Among the alternatives being examined are
 initiating subcontracting procedures early in the project planning
 phase and expediting signoffs on individual subcontracts.
   Expert systems are being examined for potential applications
 throughout the remedial program. A demonstration prototype has
 been developed  to determine tasks, resources and  costs for an
 RI/FS for landfill sites. Additional modules are expected to be
 produced and will be available to the Regions by the end of the
 next fiscal year. Another expert system is now available to estimate
 remedial construction costs for ongoing remedial assignments. This
 system was developed to assist with out-year budget projections
only.
  To facilitate  transfers of  information  about  methods  for
measuring and screening chemicals in the  field and for quick-
turnaround analyses, the U. S. EPA is developing a Catalog of
Field Screening Methods. The catalog will  be  provided to users
as both a pocket  guide  and as a disk in a dBase  III system.
Currently, the catalog includes about 30 field sampling or screening
methods including several gas chromatography methods, two x-ray
fluorescence methods, ultraviolet fluorescence, fiber optic sensors,
immunoassay, mass spectroscopy and atomic absorption.

REFERENCES
 1. COM Federal Programs Corporation. "RI/FS Improvement Analysis."
   Prepared for U. S. EPA Office of Emergency and Remedial Response.
   Washington, DC, 1987.
 2. U.  S.  EPA.  Memorandum  from  Henry  Longest.  "RI/FS
   Improvements." OSWER Directive 9355.0-20. Washington,  DC.
   July, 1987.
 3. U. S. EPA. Memorandum from J. Winston Porter. "Interim Guidance
   on Superfund  Selection of Remedy." OSWER Directive 9355.0-19.
   Washington, DC, December 24. 1986.
 4. U. S. EPA. "Data Quality Objective* For Remedial Response Activi-
   ties - Development Process." OSWER Directive 9355.0-7A. Office of
   Emergency  and  Remedial  Response, Office of Waste Programs
   Enforcement,  Office of Solid Waste  and  Emergency Response,
   Washington, DC. Center for Environmental Research Information,
   Cincinnati. OH, U.  S. EPA/540/G-87-003. 1987.
 5. U.S. EPA. "Data Quality Objectives for Remedial Response Activities
   - Example Scenario: RI/FS Activities at a Site with Contaminated Soils
   and Groundwater." Office of Emergency and  Remedial Response,
   Office of Waste  Programs Enforcement, and Office of Solid Waste
   and Emergency Response, Washington, DC. Center for Environmental
   Research  Information. Cincinnati. OH. EPA/540/G-87-004, 1987.
 6. U. S. EPA. "Compendium of Field Operations Procedures." Office
   of Emergency  and Remedial Response. OSWER Directive 9355.0-14.
   Washington, DC. 1987.
 7. U. S. EPA."RI/FS Standardized Tasks." Office of Emergency and
   Remedial Response. OSWER Directive 9242.3-7. Washington, DC,
   1987.
 8. U. S. EPA. Superfund Federal-Led Remedial Project Management
   Handbook. Office of Emergency and Remedial Response. Washing-
   ton. DC EPA/540/G-87/001. 1987.
 9. U.S. EPA. Superfund State-Led Remedial Project Management Hand-
   book. Office of Emergency and Remedial Response. Washington, D.C.
   the U. S. EPA/540/G-87/002.
10. U.  S.  EPA.  "Implementation of  the  Decentralized Contractor
   Performance Evaluation and Award Fee Process for Selected Remedial
   Program Contracts." OSWER Directive No. 9242.3-07. Washington,
   DC, 1986.
     EPA POLICY PAPERS AND GUIDELINES

-------
                        The  Potentially  Responsible  Party  Search

                                                     Donna Gerst
                                      U.S. Environmental Protection Agency
                                       John Ely, Esq.  and William Heglund
                                                    TechLaw, Inc.
                                                  Reston, VA 22091
 INTRODUCTION
  The importance of early, accurate and complete responsible PRP
 searches at sites listed on the NPL has been emphasized by Congress
 in the recent amendments to CERCLA (the Superfund Amend-
 ments and Reauthorization Act, or SARA) and by recent U.S. EPA
 statements. The U.S.  EPA's Feb. 12, 1987 policy memorandum
 entitled "Interim Guidance: Streamlining the CERCLA Settlement
 Decision Process," promotes "earlier and better" PRP searches
 to improve the timeliness and quality of information available to
 PRP  groups.
  Gene Lucero, Director of the Office of Waste Programs Enforce-
 ment, places PRP searches as the first step after identification of
 the site in his presentation entitled "The New Superfund Enforce-
 ment  Program: What Do Potentially  Responsible Parties  Do
 Now?" In addition, the U.S. EPA has updated its guidance, "The
 Potentially Responsible Party Search Manual" (Aug. 1987), which
 is intended to improve PRP searches. These developments will focus
 more  attention on the PRP search process by both the U.S. EPA
 and the PRPs. This attention,  in turn, will benefit both sides in
 the quest for fair and equitable settlements under CERCLA.

 WHAT IS A PRP  SEARCH?
  CERCLA requires those who cause waste to come to the  site
 (site operators, generators and transporters), or those who own
 the site or who did own the  site at the time of waste disposal to
 be  liable for  the cost of cleanup and for damage to natural
 resources. The PRP search is the  effort that the U.S. EPA and
 its contractors undertake in order to identify, locate and document
 those  parties who may be liable for the costs associated with Super-
 fund  sites.
  In theory, the PRP  search begins with the identification of the
 site and does not end until the site has been deleted from the NPL
 and all of the U.S. EPA and state costs: have been recovered from
 PRPs. All PRPs and information  about their waste transactions
 should be identified as early as possible.
  In practice,  for reasons of budget, timing and site complexity,
 the U.S. EPA has selected a core group of tasks to be conducted
 at every site and a group of supplemental tasks, which vary by site,
 to locate as many parties as possible and the information connecting
 them  to the site. These tasks culminate in a PRP search report,
 which recounts the history of the site, identifies PRPs and docu-
 ments their connection with the site. The investigation should iden-
tify those persons who may have further information about  the
site but are not liable or for whom information may be inconclusive.
  Until recently, the PRP search report was prepared after the  site
was proposed for or was listed on the NPL and before the reme-
dial investigation was begun. In especially complex cases, or in cases
with many potentially  responsible parties, identification  and
location of PRPs actively continued during the case.
  In the revised Potentially Responsible Party Search Manual and
the previously mentioned guidance, the U.S. EPA has outlined new
initiatives to provide for earlier and better PRP searches. Ten basic
tasks comprise the core tasks mentioned earlier - such as review
of Federal, state and local records - and are performed in every
PRP search. The manual also sets out additional tasks - such as
obtaining aerial photographs - that may be appropriate in a given
case to assure that the PRP search is as complete as possible. The
conduct of these tasks is addressed  in detail in the Manual and
is discussed in  a separate paper prepared for this symposium.

WHY "EARLIER AND BETTER" PRP SEARCHES
  In enacting SARA, Congress determined that encouraging PRP
cleanup is an efficient way to remedy uncontrolled hazardous waste
sites. Congress has given broad enforcement  authority to the
U.S. EPA and  affirmed joint and several liability while also, in
CERCLA Section 122, imposing time-limited settlement procedures
on the Agency and the PRPs. The settlement procedures include
information exchange with PRPs  and moratoria on remedial
actions to allow for negotiations. In its Superfund Comprehensive
Accomplishments Plan, the U.S. EPA has set goals,  toward
achieving settlement and enforcement in pending cases. The U.S.
EPA's policies recognize that settlement and enforcement are facili-
tated by an early and thorough PRP search. Without "earlier and
better" PRP searches, enforcement actions, and consequently, set-
tlements within the Congressionally mandated guidelines of CERC-
LA are frustrated.
  Why should this be so? Unless the U.S. EPA knows who all the
parties are, it is difficult to engage them in settlement negotiations.
The U.S. EPA almost always knows at least some of the PRPs
at NPL sites. But in the Congressional testimony underlying SARA,
the U.S. EPA was criticized for not  notifying all of the PRPs of
removal or remedial activities at a site in a timely fashion to permit
the PRPs an opportunity to undertake action. From the testimony
and from experience, the U.S. EPA believes that PRPs are more
likely to develop and implement remedies  if all PRPs have been,
or will  be, identified sufficiently early in the remedial process to
be  provided  adequate  opportunity to  become  involved.
Additionally, PRPs are more likely to perform the remedial action
if they have meaningfully participated in the remedial investigation
and feasibility study and have influenced the selection of a remedy.
  Without an adequate, timely PRP search, it is unlikely that there
will be a rapid mobilization of the remedial action,  and it is more
                                                                            EPA POLICY PAPERS AND GUIDELINES

-------
 likely that the U.S. EPA will expend Superfund moneys rather than
 relying on PRP remedies.
   Importantly, the U.S. EPA has the discretion to send "special
 notice" letters to PRPs before undertaking remedial  investi-
 gation/feasibility studies and the remedial design/remedial actions
 under this provision. These special notice letters allow windows
 for good faith negotiation of a remedy.
   Moreover, CERCLA now requires the U.S. EPA special notice
 letters to PRPs to provide names and addresses of PRPs, the
 volume and nature of waste contributed by each, and a ranking
 by volume of substances at the facility to the extent such informa-
 tion is available. The U.S. EPA is in a better position to provide
 this information when it can  identify all PRPs or other parties
 possessing such information. This information can then be more
 readily gathered for information exchange, development of volu-
 metric rankings and non-binding preliminary allocations of respon-
 sibility. CERCLA now also encourages exchange of information
 with PRPs through community relations activities and publicly
 available administrative records  of the U.S.  EPA response
 decisions.
   The PRP search is a building block for litigation in the case which
 must meet evidentiary standards. Information from the PRP search
 will become an important part of the record  should the  case be
 reviewed by a court. PRP information must be carefully docu-
 mented and referenced to support the U.S. EPA liability deter-
 minations.
   PRPs have a vested interest in seeing the searches done earlier
 and better. Once a PRP determines that it meets the standard of
 liability set out in CERCLA, its strategy shifts from avoiding lia-
 bility to reducing its costs by  assuring that the share of waste
 attributed to it is not more than it should be and determining that
 all PRPs.equitably share the cost. This strategy should encourage
 PRPs to carefully information provided by the U.S. EPA in special
 notice letters and to take steps to assure that the U.S. EPA has
 accurate information regarding all PRP involvement  at a site.
   As earlier and better PRP  searches cause earlier  and better
 information exchange,  both sides in a Superfund case will benefit.

 TRENDS IN PRP SEARCHES
  The U.S. EPA's approach to PRP searches has matured into
a strategy. The stated goal is for earlier and better PRP searches.
 The U.S. EPA has undertaken changes to bring about this goal.
 Previously, PRP searches were performed at any time prior to the
 remedial  investigation,  the U.S. EPA's  initiatives  encourage
 beginning the PRP search concurrently with proposal of the site
 to the Office of Emergency and Remedial Response for listing on
 the NPL. Additionally, the U.S. EPA has completed the first phase
 of PRP search training for Regional Project Managers and con-
 tractors. These steps will help assure that the  PRP search starts
 earlier and is complete in both identifying PRPs and in  docu-
 menting the information  that supports their listing as PRPs.
  The trend is now clearly away from viewing the initial PRP search
 report as an end in itself. The process is becoming more diffuse
 and more routinely includes such tasks as reviewing responses to
 information requests and generating volumetric rankings of PRPs.
 The trend is to tailor the PRP search strategy to the case instead
 of ordering a fast-food "PRP Search to go."
  The U.S. EPA is also implementing additional  procedures to
 assure high quality in PRP searches. The U.S. EPA is hiring a civil
 investigator for each region. Part of ehe job of the civil investi-
 gator is to develop and implement a Regional PRP search strategy
 consistent with the national goal of earlier and better  PRP searches.
 In addition, the National  Enforcement Investigations Center is
 auditing PRP searches, and the Office of Waste Programs Enforce-
 ment  is evaluating the PRP search process to assure that national
 policies  continue to promote earlier  and better searches.
  Finally, regional project managers  are calling for the assistance
 of the Office of Regional Counsel much earlier in the process so
 that the PRP search can be tailored  to the  enforcement  strategy
 in the case.
CONCLUSION
  The U.S. EPA has made a commitment to improve the timeli-
ness and completeness of PRP searches. Guidance has been issued
requiring PRP searches to be initiated quickly after site discovery.
The U.S.  EPA also is working to improve the quality of its searches
through training its contractor and regional personnel and through
evaluating the work product to assure its adequacy  Parties engaged
in PRP searches need to become familiar with the new guidance
and incorporate it into their approach to this work.
    EPA POLICY PAPERS AND GUIDELINES

-------
                                      Mass  Notification  of  Owners
                                 Under the  Defense  Environmental
                                             Restoration  Program

                                                        H.  David Hill
                                                      Real Estate Office
                                                    Timothy E. Kelleher
                                                   Gerald J.  Stadler, P.E.
                                                    Engineering Division
                                             U.S. Army  Corps of Engineers
                                                       Chicago District
                                                       Chicago, Illinois
ABSTRACT
  The Chicago  District of the U.S. Army Corps of Engineers
developed  a method to  identify and  notify owners of former
Department of Defense ammunition plants in the State of Illinois
about the Defense Environmental Restoration Program. Letters
describing  the Program and its origin, Public Law 98-212, were
mailed to approximately 585 property owners of 1003 parcels of
land comprising four former ammunition plants in diverse loca-
tions in Illinois. The former plant lands comprise almost 35,000
acres of land presently used for agricultural, commercial, indus-
trial and residential purposes.
  In the letters, the owners were provided with an 800, toll-free
telephone number which was equipped with an answering machine
and a pre-recorded message, which they were asked  to listen to,
to gain additional information about the program. They were also
asked to give any information they might have regarding the pos-
sible location of debris and hazardous or toxic waste and ordnance
and explosive waste on their land or other  land within the former
plant  site.
                  PUBLIC LAW 98-212-DEC 8. 1983
                                                  97 STAT. 1427
         Code, section 4312, may be paid subsistence and travel allowances in
         excess of the amounts provided under title  10. United States Code,
         section 4.113.
                         CLAIMS. DEJENSE

          For payment, not otherwise provided for,  of claims authorized by
         law to be paid by the Department of Defense (except for civil
         functions), including claims for damages  arising under training
         contracts with carriers, and repayment of amounts determined by
         the Secretary concerned, or officer! designated by him. to have been
         erroneously collected from military and civilian personnel of the
         Department of Defense, or from Sutes, territories, or the District of
         Columbia, or members of the  National  Guard uniU thereof;
         SI 60.400.000
                 COURT or MILITARY APPEALS.  DEPTHS!

          For salaries and expenses necessary for the United States Court of
         Military Appeals. $3.372.000. and not to exceed $1.500 can be used
         for official representation purposes.

                        SUMMER OLYMPICS

          For logistical support and personnel services (other than pay and
         nontravel related allowances of members of the Armed Forces of
         the United Sutes. except for members of the Reserve components
         thereof called or ordered to active duty to provide support for the
         XXIII Olympiad) provided by any component of the Department of
         Defense to the 1984 games of the XJCIII  Olympiad; $45.000.000:
         Provided. That the Department of Defense may also provide support
         to the Los Angeles Olympic Organizing Committee on a reimburs-
         able basis, with such reimbursements to be  credited to the current
         applicable appropriation account* of the Department.

                 ENVIRONMENTAL RESTORATION. DEPTNSE

          For expenses, not otherwise provided for. for environmental resto-
         ration programs, including hazardous waste disposal operations and
         removal of unsafe or unsightly buildings and debru of the Depart-
         ment of Defense, and including program* and operation* at site*
         formerly used by the Department of Defense; $150.000.000.

                          Figure 1

                      Public Law 98-212
  Owners were asked to call by a certain date to confirm the
currency of their ownership and to make their desires known as
to involvement in  the  program.  The 800  telephone line and
answering machine were maintained for approximately 1 year from
inception. Out-of-state owners were advised they could call another
number collect.
  This method, as opposed to contacting each owner and making
a separate survey for each ownership parcel, was an efficient means
of contacting multiple owners of a potential project site. It offered
personal contact,  convenience without cost to the  residents and
excellent documentation of their concerns. Also, the letter was
reported in local newspapers as a matter of interest and concern
to the owners.

INTRODUCTION
  During World War II, War Department ordnance plants sprang
up, seemingly overnight, on thousands of acres of Illinois farmland.
After the war, these plants were closed and the land  was  sold to
farmers, businesses and others. Sometimes, however, chemicals,
ammunition and building debris were left at the sites. To remove
such unsafe or unsightly conditions, Congress enacted Public Law
98-212,  the  Defense   Environmental Restoration  Program,
Dec. 8, 1983 (Figure 1).
  Public Law 98-212 covered  three major categories: unsafe or
unsightly buildings and debris; hazardous and toxic waste (HTW);
and ordnance and explosives waste (OEW).  Building and debris
projects involve demolition and removal of unsafe or unsightly
materials and, in  some cases, restoring the natural environment.
Ordnance  and explosives  waste  projects involve  identifying,
rendering  safe and properly disposing  ammunition, bombs,
explosives, propellants, weapon systems,  and military chemical
agents.  Hazardous  and toxic waste projects involve  identifying,
investigating, and cleaning up contaminated substances as  defined
by RCRA.
  Subsequently, the Department of Defense designated the Army
as the single  manager for environmental restoration activities at
closed installations or  "formerly-used" properties. The Army
Corps of Engineers was assigned as the executive agent for program
management and execution.
  In early 1984, geographic boundaries were assigned to the Army
Corps of Engineers' Field Operating Activities  together with a
description of the  "New  Mission Assignment."  The Chicago
District, Army Corps of Engineers was assigned as the Lead District
for the State  of Illinois, and the District was requested to  provide
an initial inventory of candidate sites within the  State by the end
of April 1984.
                                                                                 EPA POLICY PAPERS AND GUIDELINES    7

-------
  An initial inventory of 41 sites was obtained by District personnel
through disposal records of the General Services Administration
in Chicago, Illinois. The list of former installations includes, among
others, 15 NIKE Missile Battery sites near Chicago and St. Louis,
training camps, a U.S. Coast Guard Radio Station (a part of the
U.S. Navy during World War II), a 22,000 acre ordnance plant
(now part of the National Park Service's Crab Orchard Wildlife
Refuge near Carbondale, Illinois) and four other ordnance plants
comprising approximately 35,000 acres. These latter four plants,
unlike the other sites, comprised approximately 35,000 acres of
land and involved hundreds of owners and ownership  parcels.
These plants are where the method covered in this paper for iden-
tifying and notifying owners was employed.
  The Chicago District's task was to identify the owners of the
properties, inform them of the Defense Environmental Restora-
tion Program and determine if environmental problems within the
scope of the program remained at the sites.

DEFENSE ENVIRONMENTAL RESTORATION PROGRAM
PARAMETERS
  The funds which Congress provided in Public Law 98-212 were
for  environmental restoration  programs  for  environmental
problems caused  by the Department of Defense at present and
former DOD installations. (The Corps' assignment relates to former
installations.) Aside from those activities which Congress did not
intend the law to cover, such as CERCLA (SUPERFUND) and
activities under the various DOD services' Installation Restoration
Programs and the Hazardous Waste Disposal Operations of the
Defense Logistics Agency, it was necessary to establish at each site
what, if any, problems resulted from DOD activities.
  Aside from the consideration of legal obligations, such as the
case of a real estate transfer document which contained an express
warranty (an unlikely event—an express disclaimer being more
likely), a determination had to be made as to the causal relation-
ship between the existence of  an environmental problem and prior
DOD activity at a former DOD installation to provide the basis
for an administrative conclusion. Does the problem exist as a result
of prior DOD activity at the site or, for example,  activity of an
intervening owner in the chain of title? An important factor in the
determination process is whether or not beneficial use was made
of improvements in part or in whole by any intervening owners.
There are other aspects, such as the words "unsightly buildings
and debris" found in P.L. 98-212. What is "unsightly" is obviously
subject to subjective standards, requiring a resolution  on  a case-
by-case basis.
  For the Corps  to investigate former DOD installations and
process the  information culminating in a written report, each
project category [i.e. buildings and debris, hazardous and toxic
waste (HTW) and ordnance  and explosives  (OEW)] is managed
through three phases; Inventory, Engineering and Construction.
  In the Inventory phase where a Site Survey is initially made and
an initial Findings of Fact and  Determination of DOD responsibility
is made through a written Inventory  Report.
  In the Findings of Fact portion of the Inventory phase, there
are three categories of information: FORMER DOD INTEREST,
DOD USAGE and POST-DOD OWNERSHIP HISTORY. Con-
tacting the current owner/s of record to provide knowledge of the
Defense Environmental Restoration Program is necessary.

PROGRAM APPLICATION IN THE CASE
OF LARGE INSTALLATIONS
  Where former DOD installations encompass thousands of acres
and have been disposed to hundreds of owners of hundreds of
parcels, there are three choices of action. The first choice is to con-
tact each owner/s of each parcel/s by phone or other means and
prepare an  individual Inventory Report. The second choice is to
group  owners  of parcels  having  some  common  significant
characteristic, such as minimum or high probability of contamina-
tion. These owners could then be contacted and reported on as
a group. The third choice would involve grouping such common
characteristics, coupled with the simultaneous contact of all owners
of the former installation with a common  letter explaining the
program, soliciting information and a giving a toll-free telephone
number for their use. All three  choices involve availability of
records providing accurate information on each installation from
its inception through disestablishment, disposal and activities con-
ducted thereon during the intervening years to the present.
  Fortunately for most of the 41 initially inventoried sites, exten-
sive records were available to document the establishment of the
former  installations  by  the Department  of  Defense and  its
predecessor, the War Department, through  land acquisition and
construction. Such records also were available for the use and
subsequent disestablishment of the installations. Some sources of
these  records include: the General Services Administration; the
National Archives; the Military District for the U.S. Army Corps
of Engineers for Illinois (Army Installations', Louisville District,
Louisville, Kentucky; State government records;  local libraries and
historical societies; the County Assessor's Recorder's Offices and
local municipal government  records.
  Activities which have occurred at these former installations since
disestablishment are available to a lesser or greater degree at the
County and Municipal government level.  Disposal of a  whole
property in smaller ownership parcels creates an equal number of
separate sources of a chain of  recordable activities such as land
transactions, building code violations, driveway permits, construc-
tion  permits,  zoning  variations, etc. It  is necessary  to have
knowledge of the initial dividing-up of the real estate of the entire
installation at time of disposal and  the various changes (sub-
dividing, for example) through the intervening years. The record
of real estate transactions, or chain of title for each pan  of the
whole is available at  the  County Recorder's Office in Grantor-
Grantee Index or Tract Index  form, depending on the County.
  With respect to the four ordnance plants comprising some 35,000
acres, a search of the chain of title through the Recorder's Office
for each ownership parcel of these former installations would be
expensive, time-consuming and pointless, if there were no probable
cause for contamination and the owner did not want involvement.
Also, the records available from the previously mentioned sources
provide extensive and detailed descriptions of the acquisition, con-
struction, operation and disestablishment of these facilities; fairly
detailed knowledge could be obtained to divide  and group the
installation according to  probable areas of contamination.
  Accordingly, the last choice of action (to group areas by common
characteristics based on a knowledge gained from records and to
                         LIST or SITES USING

                      MASS NOTiriCATION TECHNIQUES


                            SUMMARY SHEET
     LINCOLN ORDNANCE
         DEPOT
     SPRXNGrlELO, IL.
     KANXAKZC-ELHOOD
     ORDNANCE PLANT
     JOLIET, It.
     GREEN RIVER
     ORDNANCE PLANT
     DIXON,  IL.
     SANCAHON ORDNANCE
         PLANT
     ILLIOPOLIS,  IL.
                    TOTAL ACRES
                    DISPOSED Of
4,750.1)
                 HUHBER
                Of PARCELS
                                                   OP
                      14,
-------
         LINCOLN ORDNANCE DEPOT
          SPRINGFIELD. ILLINOIS
SASCAHON ORDNANCE PLANT
• ILLIOPOLIS, ILLINOIS
                           Figure 3
                    The Four Ordnance Plants
contact the owners by sending them a letter) was selected as the
most effective in all respects for the four ordnance plants. The
plants  and their vital statistics are  shown in Figure 2. A map
showing their locations in Illinois is in Figure 3.
  Of these four plants, the Sangamon Ordnance Plant, a formerly
owned DOD installation adjoining the Town of Illiopolis, Illinois,
was selected as most representative to illustrate the implementation
of methods used to identify and notify the current owners.

DESCRIPTION OF THE SITE
  The  Sangamon Ordnance Plant was a DOD owned installation
located in the "heart" of Illinois corn and soybean country. The
plant lands, consisting of 19092.56 acres, were acquired in fee title
by the  Department of War through the Army Corps of Engineers
by purchase and declaration of taking in 1942. The reservation
lands measure approximately 4 miles east to west and 8 miles north
to south  and  are  were located in  Sangamon  County and in
Lanesville and Illiopolis Townships. Two railroads and old U.S.
Route 36 divide the land into approximately north and south halves.
The north half is in Illiopolis Township, and the south half is in
Lanesville. A map of the installation is in Figure 4.
  The  installation  was used from 1942 (during World War II)
through 1946 for the loading, assembly and packing of explosive
ordnance.  Approximately   700  structures  and  extensive
improvements were built by DOD for use in the plant's operation.
Its operation included the production of fuses, boosters, bombs,
shells,  grenades  and land mines. Explosive materials were both
refined at the plant and also brought into the plant from other
locations. Shells and other casings were brought into the plant from
metal parts production facilities and cleaned with solvents  prior
to and after loading. Major improvements included bunker and
igloo storage in support of four primary load lines, the adminis-
trative  area, staff residences, classification yards, underground
magazines, a water supply and distribution system and an exten-
sive sanitary sewer and drainage system. Typical structures at the
load lines  included change houses, assembly buildings with blast
containment walls, cleaning and painting facilities, a boiler house
                                                                     Figure 4
                                                              Sangamon Ordnance Plant
                                                                  Plant Boundary
                                         and bunkers. Two of the load lines included a TNT melt-pour
                                         building where wafers of TNT were melted prior to pouring into
                                         casings. The plant was operated by Remington Rand Corporation
                                         for the U.S. Army Ordnance Department.

                                         DISPOSAL OF THE SITE
                                           On August 25, 1945, 4,026.00 acres of fee land was transferred
                                         to the Farm Credit Administration for disposal. This acreage was
                                         in part the safety/buffer area and had remained essentially under
                                         agricultural use during  the plant operation.  These  lands were
                                         quitclaimed on a priority basis back to the owners from whom they
                                         were acquired by the Federal Farm Mortgage Corp (FFMC) under
                                         the Surplus Property Act  of 1944.
                                           On September 10, 1946, accountability for 14,947.26 acres of
                                         fee land was assumed by the War Assets Administration (WAA).
                                         The improvements/facilities of the ordnance plant were included
                                         in this acreage as well as large agricultural areas between improve-
                                         ments. The disposal deeds generally were tailored to reflect the type
                                         of facilities located on each parcel.
                                           On March 28, 1947, accountability for 120 acres of fee land com-
                                         prising the Administration Area in the north half of the  plant was
                                         assumed by the WAA. This area had been withdrawn from certifi-
                                         cation for disposal about July 1945  and  used thereafter as  an
                                         Engineer Redistribution Center.
                                           On April 7, 1950, 53.73 acres were withdrawn from General
                                         Services Administration  (GSA). This  acreage,  comprising  15
                                         ordnance contaminated sites ranging from the north to south ends
                                         of the plant, had been excepted out of previous FFMC deeds which
                                         disposed of the parcels within which they were located, in some
                                         cases, land-locked.
                                           On May 6, 1963, the 53.73 acres, as corrected, were reported
                                         excess to GSA and subsequently disposed of on a priority basis
                                                                               EPA POLICY PAPERS AND GUIDELINES     9

-------
 to the owners of the surrounding lands previously disposed of.
 These disposal deeds contained an indefinite restriction for surface
 use only and a release and "Hold and Save" of the United States.
 All deeds for these sites except  that for Parcel No.  1 were con-
 veyed in 1972,1973 and 1974 and recite that decontamination work
 was completed by the Department of  the Army on February 11,
 1970, "whereby the parcel is now safe for surface use only." Parcel
 No. 1 was conveyed in 1965 and required the Grantee to conduct
 the  decontamination  under  U.S.  Army Ordnance supervision.
 National  Security Clause restrictions  where applicable,  were
 released in the 1950s for the entire installation.

 DEVELOPMENT OF THE METHOD
   The foregoing site description and land disposals were substan-
 tially obtained from  documents through the General Services
 Administration, National Archives, Louisville District, Army Corps
 of Engineers and other sources. These records provided an accurate
 and detailed description of the acquisition and disposal of the plant
 reservation lands and the construction, operation and maintenance
 of the plant improvements.  Most of these records were ordered
 and analyzed,  summarized and copied as required in several loca-
 tions in Chicago, Illinois. These records, including audited  real
 estate acquisition and disposal  maps obtained from Louisville
 District, precisely defined the former reservation lands by Section,
 Township, Range and Meridian.
   With this information and  the cooperation and resources of the
 Sangamon  County Assessor's  Office,  Map Department  and
 Automated Data Processing Department, composite tax index maps
 were obtained for the approximately 30 sections of plant land. Each
 tax index map is a controlled aerial photograph map of two sections
 of land (1280 acres), running north  and south, with overlaid
 boundaries of ownership parcels of record and the photo date.  The
 scale is 1  in. to 100 ft. for subdivided areas and  1 in. to 400 ft.
 for farm  acreage describable by Section, Township Range  and
 Median. Each map measures 20 x 30. A miniaturized version
 without photo image is in Figure 5.
   Each parcel of land is identified with a Property Index Number
 which is also the Tax Index Number. Each Property Index Number
 (PIN) consists  of a Township number (not the Township number
 used  in the legal description), Section  number, Section quadrant
 number (100 = NW,  200 = NE,  300 = SW,  400 = SE) and  the
 lot/parcel number within the quadrant (Figure 6). The PIN is also
 the Tax Index Number which  appears on the property owners' real
 estate tax bill. As a rule, owners in rural, agricultural areas know
 their tax numbers and what each part stands for. The photographic
 aspect of each map provided a fairly comprehensive assessment
 of the ordnance plant's improvements remaining  intact.
  At  this point, the categories of the Fact Finding portion of the
 Inventory  Phase (namely. Former DOD Interest and DOD Usage)
 were  well  defined. The third category,  Post DOD Ownership
 History and contacting the owners was just starting.

 IDENTIFYING THE  OWNERS
  A  printout of all Tax Index Numbers for  ownership parcels
 within the installation boundaries was  obtained from the County
 Automated Data Processing Department. The order was by index
 number with names and addresses of owners. The computer pro-
 gram  allowed selection of blocks of numbers within whole, con-
 secutively  ordered Sections. The order of numbers as related to
 the Tax Index Maps began in  the northwest corner of the reserva-
tion and proceeded in Section numbering order to the south  end
 of the reservation. Non-applicable numbers (outside site boundary)
were discarded. The party who paid the real estate taxes was listed
as well as the owner, if the taxpayer was not the owner. This would
be the case,  for example, in  a land contract sale.

 NOTIFYING THE OWNERS
  Owners-of-record of all the tax ownership parcels for the plant
were sent a letter (Figure 7).  A map of the installation showing
                                                                                ILLIOPOLIS TWP.
                                                                             3EC3. 25  & 36 T I7H. R2W

                                                                                     Figure 5
                                                                    Tax Index Map—Miniature Without Photo Image
                                                           Section Numbers was attached to each letter (Figure 4). The letter
                                                           informed the addressees about the Defense Environmental Restora-
                                                           tion Program and the Corps responsibilities. It further asked them
                                                           to confirm their ownership of the land as identified through county
                                                           tax records as lying within the former installation and to advise
                                                           the Corps of any hazardous or toxic waste,  unsafe debris or unex-
                                                           ploded ordinance they might be aware of. Each letter address con-
                                                           tained a tax index number as  shown on the real estate tax bills.
                                                           The county ADP Department also ran off pressure address labels
                                                           which included the tax index number (one  address label for each
                                                           tax ownership  parcel). Like-owner labels  were accumulated for
                                                           mailing as an enclosure to one letter.
10
EPA POLICY PAPERS AND GUIDELINES

-------
          :.ซ  •
                          ILLIOPOLIS TWR
                          SEC. 24 TI/N! R.2W
                    BREAKDOWN OF THE PROPERTY INDEX NUMBER

      The  index numbering  system as used  by Sangamon  County,  Illinois
      involves  Township-Range  and Section  number  relationships.    This
      example is for the crosahatched land tract in the above Section 24.

      The  following  sequence  of numbers are  used: the  County  assigned
      Township number, the section number, the quarter section number (which
      remains constant.  Northwest - 100,  Northeast - 200, Southwest - 300,
      Southeast • 400), and the tract number.

      In this example, the numbers would be "09"  (Township No. - see bottom
      right of the above map), "24" (Section No. - see center of map),  "300"
      (Quarter  Section No.  - see lower right quarter and circled  number),
      "007" (Tract No.).

      The completed number would read:  09-24-300-007.  This number can now
      be used to determine the landowner/taxpayer.
                              Figure 6
                Breakdown of Property Index Number
   A toll-free telephone number with an answering machine and
pre-recorded message (Figure 8) was provided for State of Illinois
owners.  Out-of-state  owners  were  instructed  to  call  another
telephone number collect. Although a deadline in September, 1985
was set,  the toll-free telephone and answering machine remained
hooked-up for about 1 year from inception. All responses were
confirmed by cross reference to tax index number, related to the
composite tax index photographic maps and observed during visits
to the area.  The incoming  telephone calls  on the  toll-free line
(separate telephone) were processed to obtain the party's name and
address as shown on the letter address and a telephone number
and time and date when the party could be contacted. Using another
listing arranged by owner name, a cross reference to the tax number
was obtained. The process then was as  hereafter outlined.
   If the  party was immediately available, the tax index number
was obtained and the party was told to stand by for a return call.
This process precluded tying up the toll-free line for other incom-
ing calls. The procedure for answering the 800 line during business
hours,  8:00 AM  to  4:30 PM,  is shown in Figure  9.
   Prior to  returning the call, the name and tax number were
confirmed from the tax index number computer listing and the land
parcel was located on the composite tax index photographic map
by use of the tax index number and  also the former installation
plot plan by use of Section and Township numbers. The plot plan
showed the location of improvements as-built in  1942.
   During the return call, an  attempt  was  made  to answer any
questions and to  confirm  the "lay of the land" of the land parcel
as revealed through the tax index photographic maps. Discrepan-
cies between the  location of structures on the plot plan and  the
                                                                                                        DEPARTMENT OF THE ARMY
                                                                                                             23 August 1985
                                                                                                                 '
                                                                                    SUBJECT;  Sangamon Ordnance Plant - a FORMER Department of Defense
                                                                                              Installation which was located on approximately 19,000 acres of
                                                                                              land adjoining Hllopofls, Illinois, East of Springfield, Illinois
                                                                                              on Route I 72 and U.S. 36.
       Dear Property Owner:

           The Corps of Engineers is carrying out the Defense Environmental
       Restoration Program (DERP) at the above Installation, formerly owned by the
       Department of Defense (DoD). The location and boundaries of the former
       installation are shown on the enclosed map.

           The DERP is a congresslonalty authorized program (Public Law 98-212)
       that provides public funds for cleaning up and restoring sites formerly
       owned or used by DoD.  Included In the program are the removal of buildings,
       unsightly structures, or debris; ordnance wastes; and hazardous and toxic
       materials. The program only addresses conditions directly linked to Defense
       Department use of the site.

           Our fob is to identify, inventory and report on any such DoD buildings,
       structures or debris at the above Installation. Our higher command in
       conjunction with DoO  will determine if any of these items are  finally eligible
       for removal under the DERP.  In any event, the report must  show the current
       owners' desires with respect to the removal of such items located on their land.
       Removal  will no't  be considered without the owners' request.

           Accordingly, in order to learn of any such requests, and to help ensure
       that our  report will include any such buildings, structures, or debris which
       we may have previously overlooked, we are sending this letter to all owners
       of the above former installation lands as identified through the county tax
       records.

           We ask you  to call the TOLL FREE number listed bซk>w,  on or before
       13 September  1985 to confirm your  ownership of a portion of the former
       installation site.  We will discuss the program with you and tell you if it
       applies to anything on your land.  After our discussion, we would like you to
       advise us whether you want any involvement in the program.

          The enclosed map contains section township and range numbers to assist
       you In locating your property within the boundaries of the former installation.
       In addition, we have aerial composite maps that show ownership by tax Index
       numbers.

          Any Information you may have about hazardous, toxic or  ordnance
       waste resulting from former DoD activities at this installation would be
       appreciated.  Such Information will  be referred to other Corps of Engineers'
       offices with  recommendations for further Investigation and possible removal.

              TOLL FREE  TELEPHONE  NUMBER FOR YOUR USE

                              1-800-621-3813  (Illinois Only)

       Business  Hours:  Monday through Friday, 8:00 A.M. to 4:30 P.M.

       Non-Business Hours:  A telephone answering device will:**

          1.  Play a 30-second pre-recorded message.

          2.  Give you 30 seconds to give the following:

              NAME (on this letter)

             TELEPHONE NUMBER  (including area code)

             TIME AND DATE (when we may reach you by telephone)
                                     Sincerely,
                                Figure 7
             Letter to Owners of Sangamon Ordnance Plant
                              Illiopolis, IL
tax index maps were often resolved by the owner  (e.g., he had
demolished the structure in previous years). Sometimes a request
would be made to remove the remaining concrete foundation wall
and footing or slab which he had to plow around. Such remains
were usually visible on the photographic map through shadows in
rectangular formations.
  The owner was asked if he knew of any possibly hazardous or
toxic waste, ordnance  and explosives or  unsafe or  unsightly
buildings  and debris resulting from DOD activities which might
still remain on his property. If the answer is yes, confirmation as
                                                                                           EPA POLICY PAPERS AND GUIDELINES      11

-------
                        Pre-Recorded Message



  This Is  the  Chicago District,  Arn\y Corps of Engineers DERP  line.

  At the tone  repeat the first name as  It appears on your letter,

  then the area code and telephone number, and a time and date when

  we can reach you.  For example:



                           Smith,  John A,

                           217-555-1212

                           10:00  a.m., Aug 28



  If you should run over the 30  seconds, re-dial  and give the

  information  again from the beginning.  Thank you.

                            Figure 8
                      Pre-recorded Message
DISPOSITION FORM
ซ0*E -A
Frotedvrt for A^^rmi), OtiP loH-Frw Tflfpho**
Loctled in Red Ettate Office on Octl
                           Ititoratlon ปroaraซ.
           M omcซ
              *d
              *ปrto*ปel
      a. OW: Otf*MM

      e. Subject ttl***OBt Mlely dtdtcUtd to toll-frt* WMfttr. None of Uw rout i Ml wch
         it *lf*. MO*". triAt'tfi or eonftrwct calll will •en on Utlt HIM.

      c. Telephone will bt Mln- fro. 0800 to 16W Non-fr! . (bit.wtfif ซjachlซa -111 not
         ftafttfle CdlU.)

      ซ. T*lซpfeMM will to MMMd by wtd*rttfMd or HCCCO-A per tow*) fraป 0*00 to ltป.
         ป**9*tr t*wrป Will be tซUMปC*l ซ*M elfttflM pertonnel ซre my fro* UN
      t.  Tfce wtjtct telephone MM is for tncoatn^

      T.  nw Mkjtct ulcpMot tint II for r*cซlซ1nซ. cirtaU information oป1f (tuajtct to
         tucptlox btlow) to Uat •• CM call tht othtr partj beck on FTJ. and alto to
         toop fro* tylnfl 19 ow MM.
    i. *roct*Hrre for answering wbjoct telephone

      ..  *tซ rM e*tซ*r. announce  "Chicago Dlitrlct
                                        Corps Of Dtglnt*rt
      „.  Tft* orlMry oBJecUvt of ^j-^lซ| the call It to obuln lfป Mซ am tAt
         lซtt*r/t*vซlopt rtlcft Ulltr rvulntf, • ultpdOM fiM*ซr tKludfAf tfit CM*. ซn4
         • UM ซnd datt ^** •* c*ซ rttat kla/hir it that Aซbtr. M*U wrซ tปtlM*9 ซ'
         iMt wrt iHMfUt we ซcur*tซ by rvoming tmซ bปc* to cซll*rP HUM wit bt
         AMM on UtMr, or oinlop*.

      c.  *ซm*Ttr, MM ctlttrt ซy vlfH U-uy Out Uwy don't ••nt u bt lnvolvotf ซlth
         0ฃป. t*U tf accซ9Uป1t. bwt •• wit hi>* Uw UforMtlw tn ป•<•* b. ซboปซ to
         that M CM ralatt Cto call to i aarttcular pirut of Una.

      4.  If tlM Cilltr itartt U •ปป  OMItlant, politrly *tpUlit tซ*l  ^'4 ซopr*(|,tt It If
         M could ull thM tk*t.t it t*Uh KM -ป'll tM Mp*r to try W *ซปปtf iMtr
      .. Mfltt tnfonMtlofl roceUod Itolbly.

      f. If calltr lซMtinซซ niflMir/ktrMlf at i "•porter (iwwipiptr or Iv) or w tlattM
        official Md tMy don't ซ*l u bt called bซct, try to ซ•! '*0 to COM to a***,
        CiplaU ItailtitloAt of pftoM • HO ป>luitMard, uamrซrซ, or coปftr*iMaftlt. tipliln. aM try to Mt naw,
        t1tlซ, offlct, ซซOKy, ttc antf taltohOM Moftor, and that f*0 wilt call Urn Utl
        a> toon at pott loll.

      q. ttfora OMO and aftor UK), ulaซHont will ring ovt antwซrtAซ avcnlfit will be
        worklAf. - two rtd Mgntt thow u(i on MChlrtt by phone.
                              Rซat tttate laprtientatUe
                              Chicago Olitrlct
                             Figure 9
              Procedures for Answering 800 Telephone
to what might have been there during the plant's operation was
attempted with the plot plan and existence of debris with the photo
map. The owner was asked if District personnel could initially visit
his property to view such remnants and possibly make further invcs-

12    EPA POLICY PAPERS AND GUIDELINES
tigations and whether he desired involvement in the Program. He
also asked if he had made beneficial use of any of the structures
remaining.

CONCLUSIONS REGARDING THE
SANGAMON ORDNANCE PLANT
   Four categories emerged from  the third category of the Fact
Finding portion of the Inventory Phase, Post DOD Ownership
History: (1)  Land where the owners either did not respond to the
mailed letter or, where the owner initially contacted by Corps per-
sonnel indicated they did not want involvement in the Program;
(2) Lands where the owners responded affirmatively to the letter
for Program involvement where the owners initially contacted by
Corps personnel desired involvement in the Program; (3) Land
comprising  15 ordnance contaminated sites with restrictive-use
covenants  in the disposal transfer documents due to possible
remaining existence of such  contamination;  (4) Two ordnance
contaminated TNT  melt-pour buildings located on two  of the
previous IS  contaminated sites.
   Category (I) lands comprised approximately 18,000 acres of the
former DOD owned  19092.56 fee acres, and the  other three
categories comprised the approximately 1,000 acre remainder. Most
of the 18,000 acres had little or no improvements associated with
the ordnance plant and a minimum probability of the existence
of an environmental problem caused by the Department of Defense.
   Research of real estate transactions for the Category (2), (3) and
(4) lands through the County Recorder's Office indicated very little
activity for most of the parcels of record during the intervening
years since disposal. Most of the current owners were the parties
to whom the land was transferred when it was disposed of by the
United States or family heirs.

CONCLUSIONS REGARDING ALL OF
THE FOUR ORDNANCE PLANTS
   As previously stated, the Sangamon Ordnance Plant was selected
from the four ordnance plants shown in Figure 2 as being most
representative for the purpose of illustrating in this paper the imple-
mentation of methods used to identify and notify current owners
about the  Defense Environmental  Restoration Program and its
concerns.
   Actually, the methods were implemented at all four ordnance
plants more or less concurrently, and the conclusions herein recited
for the Sangamon plant were  in a general sense much the same
for the other three ordnance plants, to wit a large proportion of
each installation's acreage and its concomitant owners and owner-
ship parcels constituted,  in  a sense,  a "class-action negative"
insofar as involvement in the Defense Environmental Restoration
Program was concerned. This had been anticipated from the results
of research and  analysis of the extensive records available for each
installation (as corroborated by the photographic tax index maps)
which identified large areas of acreage containing no improvements
normally associated with environmental problems  at ordnance
plants. This  "lack of problems" was corroborated by the corres-
pondingly  large lack of responses from the owners  as identified
through the  computer listings and tax index  maps. The second
category encompassed a relatively small amount of acreage for
which the owners requested removal of debris and remnants of
buildings and structures, the existence of which was confirmed with
the photographic tax index maps  and the plot plan. The third
category was comprised of an extremely small amount of acreage
which had been identified through previous research as known con-
taminated areas (such as ordnance burn beds and TNT melt-pour
buildings), for which the owners did not desire cleanup action.

CONCLUSIONS
   Knowledge of the Defense Environmental Restoration Program,
as implemented by the U.S. Army Corps of Engineers, was con-
veyed to hundreds of owners of approximately 35,000 acres of
former Department of Defense installations through a relatively

-------
low cost and effective method described in this paper. Through       ment of their land in the program.
the use of a common letter sent to each landowner and a toll-free         The choice of combining accurate knowledge of each installation
"800" telephone number, the owners were given a no-cost means       with  a no-cost opportunity for  individual involvement by the
to voice their concerns and register their desires as to the involve-       owners proved to be the best choice.
                                                                           EPA POLICY PAPERS AND GUIDELINES    13

-------
                 Status  of  the U.S.  EPA's Pre-Remedial  Program

                                                  James R. Jowett
                                                Scott C. Fredericks
                                               Lucy Sibold, Chemist
                                     U.S. Environmental  Protection  Agency
                                 Office of Emergency and Remedial Response
                                       Hazardous Site Evaluation  Division
                                                Washington, D.C.
ABSTRACT
  This paper discusses the strategy the U.S.  EPA anticipates
employing to address the pre-remedial goals and requirements of
the SARA. It discusses SARA production goals, program opera-
tions under the current Hazard Ranking System (HRS), program
operations preparing for and following revision  of the HRS and
guidance to  be  developed to implement the strategy. It  also
describes the U.S. EPA's internal organization and the resources
available to implement the program.

INTRODUCTION
  The U.S. EPA's pre-remedial activities form the foundation for
the Superfund Remedial Program. Sites on the national inventory
must be assessed for their potential hazards to health and the
environment. Resources need to be used efficiently to identify the
priority sites requiring referral to the removal program or remedial
action. The primary building blocks in the pre-remedial program
have been Preliminary Assessments (PAs) and  Site Inspections
(Sis). The Agency has employed historically a sequential decision-
making process regarding further actions warranted at potential
NPL sites; each step provided additional information but required
more resources. The U.S. EPA tried to decide the disposition of
a site as early as possible in the assessment process.
  During the original Superfund program under CERCLA, the
Agency operated under a "sunset" assumption, believing that the
program would terminate at the end of its 5-year authorized period.
The Agency's goal was to identify and take remedial measures at
the worst 400 sites (the number set by CERCLA) from an inven-
tory of nearly 20,000. The preremedial program was geared to
screen sites. The purpose of a PA was to decide whether an SI
should be conducted and, if so, determine the SI priority. Under
this approach, many serious sites were identified as a high priority
for a SI. Some sites clearly were not a problem and were marked
for no-further-action (NFA). Unfortunately, little was known about
most sites, making  a clear decision on their  hazard  potential
difficult. These sites could not be considered as NFA. Therefore,
the sites usually were recommended as low or medium priority for
a SI.
  The SI was conducted to gather additional information to  rank
a site using the HRS (to determine eligibility for the NPL and hence
remedial dollars) and to aid in making judgments on what further
actions were warranted. Historically, a SI involved a site visit and
usually limited sample collection (e.g., 12 samples per site). Sis
ranged from  150 to 300 hr. Due to limits in resources, a backlog
of Sis developed.
  SARA establishes ambitious goals for the completion of PAs
and Sis. U.S. EPA management believes that clear decisions must
be made on sites at each step in the pre-remedial process to pre-
vent future SI backlogs. The following implementation strategy
was developed to achieve this objective.

BACKGROUND
  SARA and the accompanying conference report establish goals
and place certain new requirements on the U.S. EPA's pre-remedial
program. They include the following:

• For all sites in CERCLIS remedial inventory (the U.S. EPA's
  national inventory of potential hazardous waste sites) prior to
  the date of SARA enactment (DSE):

    complete all PAs  by Jan. 1, 1988
    complete all Sis, where appropriate, by Jan.  1, 1989
    complete HRS scoring, where appropriate, by Oct. 17,  1991

• For all sites added to CERCLIS after DSE, complete, within
  4 years, PA, SI and  HRS scoring where warranted
• Complete PAs within 1 year of receipt of a citizen petition
  (provided  for in SARA) and promptly evaluate hazards  if
  warranted
• Assure that Federal agencies conduct PAs for facilities on the
  Federal Agency Compliance Docket by Apr.  17, 1988
• Evaluate and where appropriate, place Federal facilities on the
  NPL by Apr. 17, 1989
• Revise the HRS by Oct.  1, 1988  to include on-site pathway,
  potential  air release,  food chain  route and human exposure
  factors
• Add enough sites  to bring  the NPL  to 1,600 to  2,000 by
  Jan. 1, 1991

Ability lo Meet SARA
  In early 1987, the U.S. EPA conducted a detailed analysis of
the Agency's ability to meet the goals and requirements in SARA
and the accompanying conference report.
  In performing this analysis, the Agency examined how many of
each of the pre-remedial tasks the Agency (or States under Coopera-
tive Agreements) would have to do to meet the goals and require-
ments in SARA or the conference report. CERCLIS was the
baseline on which the analysis was built and refined. CERCLIS
provided information on the number of sites as of DSE, the average
rate new sites are added and the number of PAs and Sis completed
to date. The U.S. EPA Regional Offices provided information on
historical trends for determining NFA upon completion of a PA
and an SI. CERCLIS data coupled with historical  trends for
14    EPA POLICY PAPERS AND GUIDELINES

-------
referring sites for NFA were used to determine the number of sites
that required an SI. To determine the number of sites that required
HRS scoring, the Agency assumed an NPL scoring pace that would
guarantee an NPL size that met the conference report goal.
  After establishing how many of each of these activities the U.S.
EPA (or States) had to perform, the agency examined whether,
based on existing resources, the SARA goals  and requirements
could be met. This stage of the analysis relied heavily on in-house
and contractor data, including:

• Data on the percentage of contractor hours devoted to pre-
  remedial activities versus non-pre-remedial  activities such as
  enforcement, remedial or special studies
• Amount of contractor time devoted to indirect activities such
  as management oversight and equipment maintenance.
• Number of hours required to perform PAs, Sis and to develop
  HRS scoring packages

  The U.S. EPA's analysis of SARA also relied on historical trends
in State contributions towards performing PAs and Sis as well as
anticipated State capability to perform additional work. The agency
also had to make some  assumptions in this stage of the analysis
on the number of hours needed to perform  Sis to meet the future
requirements of the revised HRS. The Agency assumed that the
new HRS would be somewhat more complex and require 1.5 times
more hours to collect data.
   From the analysis, the U.S. EPA concluded that not all of the
SARA pre-remedial goals can be  met. The U.S. EPA can:

•  Conduct PAs on all CERCLIS DSE sites by Jan. 1, 1988
•  Complete HRS scoring packages on post-DSE CERCLIS sites
   within 4 years
•  Complete PA petitions within 1 year of receipt

   The U.S. EPA may not be able to:

•  Complete Sis on CERCLIS DSE sites by Jan. 1, 1989
•  Complete HRS scoring packages on these sites by Oct. 17,1990
•  Assure that all  Federal agencies conduct  PAs by Apr. 17, 1988
•  Evaluate and list on the NPL all Federal facilities by Apr. 17,
   1989
•  Add enough sites to bring the NPL to 1,600 to 2,000 sites by
   Jan.  1, 1988

   The U.S. EPA's analysis also indicated that the HRS revisions
(and schedule) will profoundly affect the Agency's ability to address
SARA  SI and HRS goals.
Next Steps
   The U.S.  EPA's analysis identified ways to change  how it
operates the preremedial program (from its "sunset"  approach)
to significantly improve its level of SARA compliance, its ability
to accommodate HRS revisions and its effectiveness in identifying
the highest priority hazardous waste sites  that require remedial
measures.  The remainder of this paper presents  the Agency's
strategy.

STRATEGY TO MEET SARA GOALS
   The U.S. EPA will attempt to meet the SARA goals while
minimizing program disruption and addressing its limited resources
to the most hazardous sites. The U.S. EPA will change the pre-
remedial program by:

•  More effectively screening  out sites by improving PAs (more
   data, better decisions)
•  Conducting Sis more efficiently
•  Increasing the  resources available to  do PAs, Sis and HRS
   scoring packages

   This  strategy reduces the overall pre-remedial workload while
increasing resources available for the highest priority sites. It also
provides for the smoothest possible transition between the current
and revised HRS.
Near Term
Preliminary Assessments
  The U.S. EPA will conduct PAs on all CERCLIS DSE sites by
Jan. 1, 1988. In some cases, this will mean doing more PAs than
the Superfund Comprehensive Accomplishment Plan (SCAP) calls
for. The U.S. EPA will not conduct PAs on post-DSE CERCLIS
sites until it is certain that CERCLIS-DSE sites can be completed
by Jan. 1,  1988, with two exceptions: (1) PA petition sites, which
must be done within 1 year of receipt and (2) urgent situations.
Once the Agency starts PAs at post-DSE sites, they will be con-
ducted at a rate that will allow HRS scoring, if warranted, within
4 years.
  The average level of effort (LOE) for a PA will increase by 50%.
The enhanced PA will include an off-site reconnaissance and apply
standardized preliminary HRS  scoring. These additional  PA
activities will enable the Agency to make accurate and consistent
decisions about site priorities.
Preliminary Assessment Reassessment and Referral
   The Agency will reassess sites where a PA but no SI has been
done to verify that  the appropriate SI priority recommendation
has been made. In some cases, reassessment may involve off-site
reconnaissance. Because so many existing medium and high priority
sites need  Sis, the Agency probably cannot conduct Sis at all low
priority sites in the foreseeable future. The U.S. EPA will request
assistance  from the States in evaluating low priority and NFA sites.
If new information subsequently demonstrates a high priority, these
sites can be referred back to the U.S. EPA for appropriate action.

Site Inspections
   Sis started in the summer of 1987 will not be completed in time
to propose under the current HRS. The Regional Offices have two
options until more detailed information on the revised HRS and
appropriate new SI procedures can be developed:

•  Conduct Sis at medium priority sites—that  is,  those with a
   potential to score above the 28.50, the cutoff point of the cur-
   rent HRS categorized as follows:

   - sites posing near termate threat: refer to U.S. EPA's removal
    program
   - sites generating a high HRS score or involving elements that
    would be considered under the revised HRS (on-site pathway,
    potential  air release, food-chain route,  human exposure
    factors): hold for further evaluation under the revised HRS
   - sites scoring below 25.00: make NFA decision (the Agency's
    general policy is not  to rescore sites under the revised HRS
    unless new information indicates a potential problem)

   This approach will effectively screen out lower priority sites and
focus on the highest priority sites. It will also minimize the need
to revisit these sites once the revised HRS is final.

•  Focus SI activity on high priority sites—that is, those likely to
   score above the 28.50 cutoff point of the current HRS. The
   objective of this  approach is to  confirm that a  site is  a high
   priority and assess it relative to other high priority sites. Once
   the revised HRS is final, the Agency will begin a new SI approach
   at the worst sites. During the interim time period, the U.S. EPA
   will issue bulletins on likely revisions to the HRS and the impli-
   cations for SI data collection. This option will permit collecting
   some of the data likely  to be necessary under the revised HRS.
   However, the Agency will have to revisit most sites  once the
   revised  HRS is final.

HRS scoring
   Formal HRS scoring packages using the current HRS  are no
longer being developed. Only packages received before June 30,
1987 were considered for NPL Update #7 (the final update under
the current HRS). HRS scoring packages using the revised HRS
will be developed beginning in April 1988.
                                                                              EPA POLICY PAPERS AND GUIDELINES    15

-------
Federal Facilities
  The U.S. EPA has established a Federal Facilities Task Force
in the Office of Waste Programs Enforcement (OWPE) to focus
on issues relating to CERCLA/SARA as well as to RCRA. The
U.S. EPA has informed Federal agencies that they must perform
PAs and Sis; information required is equivalent to those described
in the NCP. To give Federal agencies time to compile the neces-
sary information, the U.S. EPA intends to propose a separate NPL
update under the current MRS for Federal facility sites in  early
1988.

FIT Resources
  The Field Investigation Team (FIT) contracts provide the primary
investigative capability of the  Agency's pre-remedial  program.
Their main purpose is to conduct PAs and Sis. Since 1980,  these
contracts  have  provided dedicated,  multidisciplinary  teams of
professionals and para professionals in  each  U.S. EPA Region.
No other existing or planned contracts can provide  this level of
dedicated supoort to the Regions for this work.
  The two current FIT Zone contracts are straight 5-year contracts
with options to increase LOE hours by SO^o from the initial base
of approximately 800,000 hr. The Zone I FIT contract (U.S. EPA
Regions 1-4) began with approximately  200 dedicated  personnel
Zone II (Regions 5-10) with approximately 220 persons in the pro-
gram. To make more FIT resources available to implement the
near-term strategy, U.S. EPA  is:

• Working with the FIT  Zone Program Managers to increase
  efficiencies in the areas of training, quality assurance support,
  general technical assistance, equipment calibration and program
  management.
• Increasing FIT LOE by 50"% over the  base, providing 630 full-
  time dedicated personnel nationwide. In FY 88, the two con-
  tracts are expected to provide approximately 1,257,000 direct
  labor hours.  U.S. EPA is also minimizing  FIT support of
  activities outside the pre-remedial  program. The FIT contract
  must now be primarily used to conduct PA,  SI and HRS related
  activities only, which is  a departure from the past practice of
  actively supporting other  remedial programs with FIT resources.
  In some Regions, other program support has accounted for over
  20% of the FIT LOE. In light of the  substantial pre-remedial
  workload imposed by SARA, this level of program support can
  no  longer continue.

  The U.S. EPA's near-term strategy does not involve doing a sub-
stantially greater number  of PAs and  Sis. Rather, it involves
improving the quality of these efforts using FIT LOE hours made
available through efficiency  measures.

Long Term
  The transition between the near and long-term strategy will occur
when U.S. EPA knows enough about the revised HRS to affect
the way it conducts Sis. This should happen in  early 1988, after
the revised HRS has been proposed and the nature of comments
can be ascertained. The long-term strategy is depicted in Figure 1.
     •a
   ruHTHin
    ACTION
                           Figure 1
             Proposed Pre-Remedial Program Strategy
Preliminary Assessments
  PAs on post-DSE CERCLIS sites are likely to begin in early
1988. PAs will address any additional data elements of the revised
HRS. These enhanced PAs will be conducted at a rate that allows
HRS scoring, if warranted, within 4 years. PA petition sites con-
tinue to be conducted within 1 year of receipt. PAs will result in
one of three SI recommendations:

• NFA—those sites with no potential to score above the HRS
  cutoff. They will be referred to the States
• Medium priority—those sites with a potential to score above the
  HRS cut-off
• High priority—those sites likely to score above the HRS cutoff

Site Inspections
  The  SI process under the revised HRS will assign sites to two
levels:

• Screening SI. On most sites, the Screening SI would:

    collect additional data to prepare a more refined preliminary
    HRS score
    establish priorities among sites most likely to qualify for the
    NPL
    identify the most critical data requirements for a Listing SI.
    A  Screening SI will not have rigorous data quality objectives
    (DQOs) Based on the refined preliminary HRS score, the site
    will then be referred to the State, another applicable Federal
    program or to The Listing SI level.

• Listing SI.  Sites most likely to qualify for the NPL are candi-
  date for a Listing SI. It would address  all the data requirements
  of the revised HRS using NPL-type DQOs.
  The  two-level  SI process is intended to  ensure that limited
resources are  applied  only to the highest priority sites bound for
the NPL.

Federal Facilities
  PAs or Sis received after September  1987 are not likely to be
evaluated under the current HRS. Instead, they will be held (as
in the case of non-Federal sites) for evaluation under the revised
HRS. The OWPE Task Force will coordinate all Federal facility
CERCLA and RCRA activities as well as  develop  policy and
guidance.

State Resources
  Additional funds for State Multiple Site Cooperative Agreements
may be made available. These funds could be used to conduct more
PAs and Sis. The U.S. EPA will provide additional funds only
to States with procedures in place to produce quality PAs and Sis.

FUTURE GUIDANCE
  To fully implement this SARA strategy, the Agency plans to issue
detailed guidance on the  following pre-remedial topics:

• PA guidance. Specific guidance on the enhanced PA addressing
  the process, content and recommendations. Requirements for
  off-site reconnaissance, a preliminary HRS scoring procedure,
  and revised HRS data elements will be included, as will PA proce-
  dures for reassessing existing PAs.
• Criteria for categorizing pre-remedial sites as NFA, medium and
  high priority Sis after performing a PA and a Screening SI
• PA petitions. How to track and administer PA petitions and
  interpret regulations governing petitions; also a brochure to the
  public on how to petition the U.S. EPA for a PA.
• Site Screening Analysis. An automated model and users manual
  for developing preliminary HRS score ranges after the PA and
  Screening SI
• Revised HRS/SI. Interim advisories on potential changes in SI
  data collection requirements as a result of probable revisions to
  the HRS
• SI guidance. Final  Screening and Listing SI data collection
16    EPA POLICY PAPERS AND GUIDELINES

-------
  requirements and procedures under the revised HRS
• Revised HRS scoring packages. Guidance on how to assemble
  scoring  packages

ORGANIZATION
  The U.S. EPA Headquarter's pre-remedial program was re-
organized in the spring of 1987. Previously, one branch was respon-
sible for PAs, Sis, HRS packages and the NPL. Currently, two
branches are responsible for this work:

• Hazard Ranking and Listing Branch.
• Site Assessment Branch (consisting of the Site Evaluation and
  Guidance Section and Site Operations and Contracts Section).

  The Site Evaluation and Guidance Section is developing the new
guidance on PAs and Sis, as well as  the implementation strategy
for the pre-remedial program. These  revisions to the pre-remedial
program impact  the resources needed, the technical approach for
field  activities and the decision criteria employed in the evalua-
tion process. The Operations Section will work with the Regions
to implement this effort.

CONCLUSION
  SARA specifically addresses many aspects of the pre-remedial
program. Both Congress and U.S. EPA management have clearly
delineated the importance of a solid foundation for the Superfund
Program. The reorganization of the U.S.  EPA's pre-remedial
program provides a focal point for the development of guidance
for specific activities and an overall implementation strategy. The
Agency is striving to use its contract resources effectively and to
help States develop strong programs to address SARA's goals. This
paper has discussed the fundamental concepts  the U.S. EPA
believes are essential to establishing a quality pre-remedial program.
Changes in the basic pre-remedial evaluation activities (PAs, Sis,
HRS) are being coupled with clear decision criteria. As a result,
the U.S. EPA believes the pre-remedial program will make better
quality  and sounder decisions.
                                                                               EPA POLICY PAPERS AND GUIDELINES
                                                            17

-------
                                       Meeting  the  New  OSHA
                        Hazardous Waste  Training Requirements

                                             Martin G. Kemplin,  CIH
                                           Woodward-Clyde Consultants
                                                 Wayne, New Jersey
                                                Philip L. Jones, CIH
                                           Woodward-Clyde Consultants
                                         Plymouth Meeting, Pennsylvania
 ABSTRACT
   A hazardous waste consulting firm (Woodward-Clyde Con-
 sultants) has implemented a training program to comply with
 29CFR 1910.120 (Federal Register Dec. 19, 1986). These OSHA
 hazardous waste training requirements are part of a general safety
 and health program for hazardous waste site workers.
   All employees "exposed to hazardous substances, health hazards
 or safety hazards" receive a minimum 40 hr of initial training (U.S.
 EPA Level C Training). On-site management and supervisors
 receive at least 8 additional hr of specialized training. Ail employees
 and managers receive 8  hr of refresher training annually.  Ex-
 perienced trainers are used for each level of training.

 INTRODUCTION
   President Ronald Reagan signed the Superfund Amendments
 and Reauthorization Act (SARA) on Oct. 17,  1986. This  Act
 required OSHA to promulgate regulations to protect hazardous
 waste site workers by Dec. 16, 1986. An interim final rule was pub-
 lished by OSHA in the Dec. 19,  1986 Federal Register. The  new
 regulations under 29CFR 1910.120 include comprehensive require-
 ments for worker medical surveillance, training, exposure moni-
 toring, site safety planning,  decontamination, site control  and
 emergency response. Section (e) of the regulations (below) addresses
 training requirements.

"(e) Training.
     (1)  All employees (such as equipment operators and general
         laborers)  exposed  to  hazardous substances,
         health  hazards or  safety  hazards shall  be
         thoroughly trained  in the following:

         (i)   Names of personnel and alternates respon-
             sible for site safety and health;

         (ii)  Safety, health and other hazards present on
             the site;

         (iii) Use of PPE;

         (iv) Work practices by which the employee can
             minimize risks from hazards;

         (v)  Safe use of engineering controls and equip-
             ment on the site;

         (vi) Medical surveillance requirements including
             recognition of symptoms and signs which
             might indicate over exposure to hazards;
                                                                 (vii) Paragraphs (G) through (K) of the site safety
                                                                     and health  plan  set  forth  in  paragraph
                                                                     (i)(2)(i) of this section.

                                                             (2)  All employees shall at the time of job assignment
                                                                 receive a minimum of 40 hr of initial instruction
                                                                 off the site and a minimum of 3 days of actual
                                                                 field experience under the direct supervision of
                                                                 a trained, experienced supervisor. Workers who
                                                                 may be exposed to unique or special hazards shall
                                                                 be provided additional training.  The level of
                                                                 training  provided  shall be consistent with the
                                                                 employee's job function and responsibilities.

                                                             (3)  On-site management and  supervisors directly
                                                                 responsible  for  or who supervise  employees
                                                                 engaged in  hazardous waste operations shall
                                                                 receive training as provided in paragraph (e)(l)
                                                                 and (eX2) of this section and at least 8 addition-
                                                                 al hours of specialized training on managing such
                                                                 operations at the time of job assignments.

                                                             (4)  Trainers shall have received a level of training
                                                                 higher than and including the subject matter of
                                                                 the level of instruction that they are providing.

                                                             (5)  Employees shall not participate in field activi-
                                                                 ties until they have been trained to a level re-
                                                                 quired by their job function and responsibility.

                                                             (6)  Employees and supervisors that have received
                                                                 and successfully completed the training and field
                                                                 experience specified in  paragraphs (e)(l), (eX2)
                                                                 and (e)(3) of this section shall be certified by their
                                                                 instructor as having completed the necessary
                                                                 training.  Any person who has  not been so certi-
                                                                 fied or meets the  requirements of paragraph
                                                                 (e)(l) of this section shall be prohibited from
                                                                 engaging in  hazardous  waste operations after
                                                                 Mar. 16, 1987.

                                                             (7)  Employees who are responsible for responding
                                                                 to hazardous emergency situations that may
                                                                 expose them to hazardous  substances shall be
                                                                 trained in how to respond to expected emer-
                                                                 gencies.

                                                             (8)  Employees specified in  paragraph (e)(l) and
18
EPA POLICY PAPERS AND GUIDELINES

-------
         managers specified in paragraph (e)(3) of this
         section shall receive 8 hr of refresher training
         annually on the items specified in paragraph
         (e)(l) of this section and other relevant topics.

     (9)  Employers who can show by an employee's work
         experience and/or training that the employee has
         had initial training equivalent to that training re-
         quired in paragraphs (e)(l), (e)(2) and (e)(3) of
         this section shall be considered as  meeting the
         initial training requirements of those paragraphs.
         Equivalent training includes  the training that
         existing employees might have already received
         from actual, on-site experience."
COMPLIANCE
  Woodward-Clyde Consultants (WCC) is a multi-disciplinary
hazardous waste consulting firm with offices in principal cities
throughout the United States and Europe. The firm currently has
approximately 300 employees actively engaged in hazardous waste-
related operations. Health and safety functions within the firm are
coordinated by a Corporate  Health and Safety Administrator
(CHSA) who reports to the Vice President  for Pratice (responsi-
ble for quality control, professional development and health and
safety). The CHSA is assisted by Group Corporate Health and
Safety Officers (CHSOs) who cover the Western,  Central and
Eastern Operating Groups. Each office within these groups has
its own designated  Health and  Safety Officer (HSO).  Each
hazardous waste project is assigned at least one Site Safety Officer
(SSO).
  Because of its previous involvement in hazardous  waste opera-
tions, WCC had already implemented many of the requirements
of 29CFR 1910.120 prior to the publication of the  interim final
rule. Policies already in place included an ongoing medical moni-
toring program, training, use of  site safety plans, emergency
response planning, etc. A thorough review  was made of the new
regulations and steps were implemented to assure total compliance.
  WCC already appreciated the importance of proper training for
hazardous waste site workers. As a matter of Corporate policy,
such training  has been required for all employees working on
hazardous waste work prior to the OSHA regulations.
  The new OSHA regulation is somewhat vague in terms of who
actually is covered. It is clear that CERCLA and RCRA sites are
covered as are those sites that have been designated for cleanup
by State or Local Governmental Authorities. Emergency response
operations and  post-emergency response  operations also are
covered.
  Hazardous waste work for private clients at non-CERCLA or
non-RCRA sites apparently is not  covered. WCC has chosen as
a matter of policy to apply these regulatory requirements to its em-
ployees at such sites.  Municipal or sanitary landfills handling
domestic wastes or operations which do not involve exposure to
hazardous substances (such as work in a clean zone) are not covered
by these regulations.
  WCC has structured their training program both to comply with
the OSHA regulations and U.S. EPA terminology. Accordingly,
WCC calls their 40-hr introductory course a "Level C" course
because it emphasizes the use of air-purifying  respirators (EPA
Level  C). The 16-hr advanced course  for  site  managers and
specialized employees is commonly referred to as the "Level B"
course because the emphasis is on the use of supplied air respira-
tors (airlines and SCBAs) and advanced emergency response proce-
dures and exercises. The U.S. EPA refers to working in  supplied
air respiratory equipment as Level B protection. U.S. EPA Level
A work involves using both supplied air and  a totally encapsulating
suit. WCC does not certify personnel for U.S. EPA Level A work.
Level A protection is used primarily for transportation emergency
response type work.
Introductory Training Program Content
  The WCC 40-hr introductory (initial instruction) off-site training
course is run as 4 long days (8:00 AM — 5:30 PM) with assigned
evening readings to achieve the 40-hr time requirement (average
of 10 hr/day). A typical agenda is as follows:

DAY ONE

  Agenda/Program Overview
  Types of Hazards and Hazard Recognition
  Basics of Toxicology
  Exposure Limits
  Threshold Limit Values
  Finding Chemical Safety Information
  Introduction to Respiratory Protection
  U.S. EPA Levels of Protection
  Medical Surveillance Program
  WCC Hazardous Waste Health and Safety Program

DAY TWO
• OSHA Regulations for Hazardous Waste Workers
• Heat  Stress and  Cold Stress
• Personal Protective Equipment (PPE) With Emphasis on Skin
  and Eye Protection
• Small Group Sessions (Groups Rotate for the Balance of the Day
  through the following Areas for Hands-on Experience):
  -  Group I—SCBA/Airlines
     Group II—Respirator Fit-Testing
  -  Group III—Monitoring Instruments (OVA, HNU, CGI, etc.)

DAY THREE
  Decontamination Procedures (Personnel and Heavy Equipment)
  Site Operations and Procedures
  Contingency Problem Set (Emergency Response Review)
  Subcontractor Health and Safety
  Specific Site Case Studies
  Site Safety Plans
  Site Safety Plan Development Workshop (attendees are split into
  groups of five to  six persons per group. A project manager and
  SSO is designated for each group. A simulated site is presented
  (the field exercise site) and each group must prepare and receive
  proper approvals for a  site specific safety plan).

DAY FOUR
• Site Safety Briefings (SSO reviews the Site Safety Plan with their
  group and  has each member sign the acknowledgement page
  stating that they have read the plan, understand it and agree to
  abide by it).
  Fire Extinguisher Use
  Level C Task (Site Mapping and Soil Sampling)
  Level B Task (Confined Area Entry)
  Examination (Written)
  Field Exercise Review
  Course Evaluation Form Completion
  Course Review/Questions

  All attendees at the Level C course are required to be clean shaven
and to have a current medical approval  to wear an air-purifying
respirator. They are encouraged to bring rubber boots to the course
to permit full participation in the field decontamination exercise.
Dress for the course is casual and a relaxed, open atmosphere is
encouraged to make attendees feel at ease.
  Most  presentations are accompanied by extensive slide audio-
visual aids. WCC has a training slide library of over 1000 color
slides for this purpose. Each trainer is an experienced health and
safety professional with extensive hazardous waste site experience.
Certified Industrial Hygienists (CIH), safety professionals and toxi-
cologists present the training lectures.
  Field exercises are made as realistic as possible with full donning
of PPE, monitoring instrument calibration and use and actual site

            EPA  POLICY PAPERS AND GUIDELINES    19

-------
sampling and mapping. Emergency response scenarios are staged
and rescue operations are performed. Each group  must obtain
accurate information on local emergency response services prior
to starting the field exercise (fire, police, ambulance telephone
numbers,  route to the closest hospital with an emergency room,
etc.).  Emphasis is placed on proper  site communications using
walkie-talkies and hand signals. The "Buddy System" is maintained
at all times. Project Managers are responsible for presenting a field
report on the group's findings at the end of the field exercises.

Advanced Training Program Content
  The attendees at the "Level B" course are supervisory person-
nel and individuals who are expected  to be assigned to more ad-
vanced waste site activities. They must be in good physical condition
and clean shaven  to attend the course. A sample course agenda
is as follows:

DAY ONE
• Review of Respiratory Protection (Emphasis on  Airlines and
  SCBAs
  Waste Site Management Problems
  Emergency Response Procedures
  Confined Space Entry Procedures
  Level B Logistical Problems
  Small Group Sessions on SCBAs, Airlines
  Level B Case Studies
  Level B Safety  Plan Development

DAY TWO
• Site Safety Briefing
• Field Exercise  (Including Emergency Response  and Rescue
  Exercises)
• Decontamination
• Debriefing
• Course  Review
• Written Examination
• Course  Evaluation

Annual Refresher Training
  Each hazardous waste employee receives an 8-hr annual refresher
training course in health and safety. These courses are scheduled
at individual offices on a rotating basis to catch employees between
field assignments.  Refresher content includes the following:

  Repeat Fit-testing with an Air-Purifying  Respirator
  Review  of the Medical Surveillance Program
  Review  of Incidents and Case Studies
  OSHA Hazardous Waste Worker Protection Regulations
  WCC Health and Safety Organization
  Site Safety Planning
                                                                  • Monitoring Instrument Calibration and Use
                                                                  • Discussion of Site Hazards
                                                                  • Questions and Answers
                                                                  DISCUSSION
                                                                    All employees who have completed initial training receive 3 days
                                                                  of actual  field  experience under  the  supervision of a trained,
                                                                  experienced supervisor. A conscious effort  is made to  select
                                                                  employees for the most difficult jobs who have the highest degree
                                                                  of expertise and experience. Employees who serve as SSOs are to
                                                                  be experienced  site workers with current first aid  training.
                                                                    The optimum group size for a good training program is 20 or
                                                                  fewer attendees. Group sizes over 20 tend to be difficult to manage.
                                                                  Larger groups also conflict with a more relaxed atmosphere and
                                                                  good "give-and-take" during question and answer periods. The
                                                                  larger groups also require more complicated logistics in terms of
                                                                  providing adequate PPE and  monitoring instrumentation.
                                                                    The class is broken up into smaller groups for  the "rotating
                                                                  sessions" portion  of  the 40-hr course.  As an example, Group A
                                                                  will be fit-tested with respirators while Group B is receiving in-
                                                                  struction in monitoring instrumentation and Group C is receiving
                                                                  an introduction to SCBAs. The groups rotate to a new area on
                                                                  a predetermined schedule. Several training  rooms must be avail-
                                                                  able  for this portion  of the course.
                                                                    Attendees are encouraged to ask questions during  presentations.
                                                                  Each attendee is provided with a training manual  specially pre-
                                                                  pared for that particular type of training course. A truthful depic-
                                                                  tion of potential hazards on waste sites is presented as clearly as
                                                                  possible. The emphasis is placed on hands-on  experience with equip-
                                                                  ment and procedures. Students are told that they will be given a
                                                                  written examination at the end of the course and asked to fill out
                                                                  a course evaluation form. Sign-in sheets are used to each record
                                                                  attendee's  presence, affiliation and social security number.
                                                                    Upon successful completion of a course, each attendee receives
                                                                  a signed health and safety training certificate. Each attendee also
                                                                  receives a  laminated photo  ID card  for  the  level  of training
                                                                  achieved.
                                                                    Documentation  of training  is kept as follows:

                                                                    Sign-in Sheets
                                                                    Completed Fit-test Forms
                                                                    Examination  Papers
                                                                    Site Safety Plans and Reports from the  Field Exercise
                                                                    Course Evaluation Forms

                                                                  CONCLUSION
                                                                    WCC has attempted to refine a health and safety training pro-
                                                                  gram which not  only meets the requirements of 29CFR 1910.120,
                                                                  but also provides valuable and workable employee protection pro-
                                                                  cedures for activities  on diverse types of hazardous waste sites.
20
      EPA POLICY PAPERS AND GUIDELINES

-------
              Potentially Responsible  Party  Search  Methodologies

                                                  Donna Lee Gerst
                                     U.S. Environmental  Protection Agency
                                     Office of Waste  Programs Enforcement
                                                 Washington, D.C.
                                                 Laurie A.  Redeker
                                     PRC Environmental Management,  Inc.
                                                   Chicago,  Illinois
ABSTRACT
  One of the most important aspects of CERCLA enforcement
actions is to determine which parties are responsible for contami-
nation at a site and whether those parties can pay for cleanup. These
determinations generally are made through PRP searches. The U.S.
EPA, with the assistance of Technical Enforcement Support (TES
II) contractors, has developed a manual on conducting PRP
searches. The purpose of the manual is to provide uniform guidance
to U.S. EPA,  state  and  contractor personnel involved in PRP
searches and to U.S. EPA and state personnel who incorporate
PRP searches into an enforcement strategy.  The manual is par-
ticularly significant  since current U.S.  EPA policy and SARA
encourage PRP searches early in the Superfund cleanup process.
This paper provides an overview of the manual and how it can
be used.

INTRODUCTION
  During the Superfund cleanup process, a PRP search is con-
ducted to identify parties liable for cleanup at a site. A preliminary
PRP search is conducted at the time of site discovery to determine
obvious PRPs who may be available  to finance removal actions
at the site.
  Once a site  is submitted for inclusion on the NPL, a more
extensive baseline PRP search  is conducted. Each baseline PRP
search consists of 10  basic tasks described in the manual.  In
addition to the 10 tasks that constitute the baseline search, a PRP
search usually includes one or more specialized tasks described in
the manual as deemed necessary by the regional project manager
and as dictated by the complexity of the site. As site cleanup
progresses, further tasks may be initiated to help identify all PRPs
for a site as well as their financial status and other pertinent data.
The manual describes 18 specialized tasks that can supplement the
baseline search.
  In all, the manual describes 28 tasks. Each task description in
the manual includes objectives, procedures,  and  problems/reso-
lutions in a concise, easy-to-read format that can be used as a check-
list to indicate which steps and tasks  have been completed. This
paper will: (1) demonstrate how the manual can be used to conduct
a PRP search and (2) briefly describe the 28 PRP search tasks.

MANUAL ORGANIZATION
  The manual is composed  of two  parts:  background and
methodology. The background section  defines a PRP and discusses
the role of PRP searches under CERCLA (42 USC 9601), while
the methodology section provides detailed descriptions of different
tasks performed in PRP searches. Nine appendices provide sample
PRP search reports as well as information sources, relevant policy
and guidance documents, an activities checklist and a glossary for
easy reference.
  The methodology part of the manual is divided into two sections:
the 10 basic tasks and the 18 specialized tasks. The section on the
18 specialized tasks  is divided  into 4 subsections:  specialized
information sources, waste stream comparisons, databases and
other tasks.
  Each task is covered in one to six pages depending on the com-
plexity of the task.  Each task  is divided into three sections:
objective,  procedures and problems/resolution. The objective
section may include what the task is to accomplish, when the task
is to be used and references to sample reports in the appendix that
contain the task.
  The procedures  section is divided into  "Initial Information
Needs" and "Process." The initial information needs section points
out what background information the researcher will need to begin
the task. For example, before  conducting a title search, the
researcher needs to locate the site on a map and obtain a legal
description and determine the time period the title search is to cover,
whether certified copies of any of the title documents are required
and the type of documents to copy. The process section outlines
the steps to complete the task including information sources and
the types of information available.
  The problems/resolution section  identifies  problems that are
commonly encountered while completing the  task. At least one
resolution is presented for each  problem.
  Throughout the methodology section there are cross-references
to other tasks. For example, the interviewing tasks reference the
file review task because names of interviewees  often are obtained
from reviewed files. An inspection report from the period during
which a site was active may list the inspector's  name. The inspec-
tor may have additional site information not recorded elsewhere.
  Nine appendices have been included to assist the PRP researcher.
These appendices are listed below.
• Appendix A,  Glossary and Acronyms—defines many of the
  terms used in the manual. These terms  are presented in bold
  face type throughout the manual.
• Appendix B, Activities Checklist—presents each of the 28 tasks
  along with a level-of-effort estimate and selection criteria. This
  checklist serves as a menu of the possible activities that can be
  selected during a search. This can assist the researcher who is
  unfamiliar with: (1) the number of hours necessary to complete
  a task or (2) when a particular task would be useful.
• Appendix C, U.S. EPA/NEIC Information Services—explains
  the relevant information NEIC can provide to U.S. EPA Super-

           EPA POLICY PAPERS AND GUIDELINES    21

-------
   fund personnel.
 •  Appendix D, Key Information Source Index—lists document
   sources and document types that are useful in obtaining infor-
   mation. Sources are provided for each task. Knowing the types
   of documents and the types of information available from these
   documents can assist the researcher during  files reviews.
 •  Appendix E, Information Collection  Forms—provides sample
   forms used to conduct interviews and industrial surveys. These
   forms must be altered to fit site-specific needs; however, they
   serve as an outline of the types of information that may be col-
   lected during these tasks.
 •  Appendix F, CERCLA—is a copy of the act as amended in 1986
   by SARA.
 •  Appendix G, Guidance and Policy Memoranda—contains copies
   of several memoranda issued by U.S.  EPA, many of which are
   being revised under SARA. These memoranda provide guidance
   for  various PRP search  issues. New and  revised guidances
   relating to PRP searches will be included in the manual as they
   become available.
 •  Appendix H, List of Contacts—provides names, addresses and
   telephone numbers of federal and state agencies that may be con-
   tacted to obtain  site-specific information.
 •  Appendix I, Sample Reports—provides several sample work
   products. None of the reports are presented in their entirety;
   only highlights are presented. These sample reports can be used
   as guides for standard format, and they are examples of the type
   of information that  should  be included  in PRP reports.

   It is anticipated that the manual can step the PRP researcher
 through the necessary procedures to successfully complete a PRP
 search for any type of site. Because every detail and anticipated
 problem cannot be included in a manageable document, the manual
 attempts to present enough information to  get the research start-
 ed at each step of the  search in an easy-to-use format.

 BASIC PRP SEARCH TASKS
   The 10 basic PRP search tasks are, in the order they are gener-
 ally performed:

   Agency record collection and file review
   Title search
   Interviews with government officials
   CERCLA 104(e)/RCRA 3007 letters
   Records compilation
   History of operations at the site
   PRP Name and address update
   PRP Status/PRP history
   Financial status
   Report preparation

   The first three tasks—government agency record collection and
 file reviews, title  searches  and interviews with public officials—
 are used to identify  PRPs.  In most cases, these three tasks can be
 conducted simultaneously.

 Agency Record Collection and File Review
   The objective of the agency record collection and file review task
 is  to locate and review all government records pertinent to the site
 and the PRP search. Relevant records could include correspon-
 dence, hazardous waste manifests, technical data,  permits and
 complaints. These records provide important  information necessary
 to become  familiar with the site, identify  PRPs and determine
 additional sources of PRP information.

Title Search
  The title search is  used to identify past and present site owners.
Past owners are liable if they were owners at the time of disposal.
Other PRPs may also be identified if their roles appear in recorded
documents. Specific results of the title search can include summaries
of transactions involving the site property, copies of title docu-
ments identifying past and present owners and possibly lease agree-
ments with site operators. A title search may be conducted for
parcels of land adjacent to the site. This may provide names of
people who are familiar with past or present site activities; these
people can be contacted and interviewed. A title search for adjacent
parcels may also provide information about other activities in the
area that may have contributed to site contamination.
Interviews with Government Officials
  Interviews  with government officials are conducted  simul-
taneously with file reviews.  The objectives of this tasks are to:
(1) identify government agencies that may have relevant documents
or information and (2) develop information on site operations, site
history and PRPs. This can be a very productive research avenue
because government officials, especially state and local officials,
often work directly with hazardous waste sites and  have more
intimate knowledge of the sites. The interviews may reveal valuable
personal recollections not  recorded in documents.

CERCLA 104(e) and  RCRA 3007  Letters
  A fourth  task  used to  collect information is the issuing of
CERCLA 104(e) or RCRA 3007 letters. The  objective of issuing
these letters is to formally request information about hazardous
waste management practices at the site from persons familiar with
the  site. The names of persons initially sent letters usually are
identified during the three tasks described earlier.
  Information commonly requested includes information such as
that concerning site operations, chemical usage and storage, waste
generation  and  waste disposal.  Examples  of documentation
provided by PRPs or other persons may include hazardous waste
manifests, site maps, purchase orders, weight tickets, technical data
and permits.  This documentation is screened for relevant, site-
related information. It may  identify new PRPs  or  other parties
who can then  be contacted for additional information.
  After the researcher begins collecting the initial information on
site activities and the PRPs, the remaining  six basic tasks are
initiated.

Records Compilation
  Records compilation is started after the agency  files are located.
The objective of this task is to organize the files into a useful and
easily accessible source of information, as well as maintain accurate
documentation of all findings. This records organization should
allow easy retrieval of information by providing an index to access
the  information desired.

History of Operations at the Site
  After obtaining information about site, researcher develops a
site history  and a list of PRPs. During the history of site opera-
tions task, the researcher uses the previously collected informa-
tion to develop a history of the site. This history  generally begins
immediately preceding the first industrial use  of  the site and
continues to the present. It includes  site-specific information con-
cerning waste  generation and transportation to the site,  waste
disposal methods and  environmental enforcement actions. After
developing the history of site operations, a list of PRPs and their
involvement with the site can be formulated.
  Once a list of PRPs is formulated, additional information is
obtained on each PRP. This  information includes current names
and addresses, corporate  or personal  status and  history and
financial status.

PRP Name and Address Update
  Updating the PRPs  names and addresses is often  fairly simple
but important task. Current names and addresses are required for
the U.S. EPA to send CERCLA 104{e) or RCRA 3007 letters. Other
information, including current corporate address,  registered  agent,
mergers, name changes and dissolution, may be collected to ensure
that all PRPs are contacted.

PRP Status/PRP History
  The PRP status/PRP history task is used in conjunction with
the PRP name and address update task. The objective of this task
22    EPA POLICY PAPERS AND GUIDELINES

-------
is to develop background information on companies or individuals
identified as PRPs, thus the enabling U.S. EPA to carry out addi-
tional tasks such as corresponding with PRPs  and conducting
financial assessments. Corporate information should include the
date of incorporation, whether PRPs currently exist, the fate of
inactive companies and names of parent or successor companies.
Information obtained for individuals and unincorporated compa-
nies should include their  current location, their association with
other PRPs (such as a company officer for another PRP) and their
involvement with the site.
Financial Status
  The financial status task is completed once PRPs are identified
to obtain some indication of a company's or individual's ability
to pay for remedial action. The completeness of the financial in-
formation depends on the amount of public information availa-
ble. This type of financial information is available through sources
such a Dun ! Bradstreet reports, local tax assessor offices,  Secre-
tary of State offices and the Securities and Exchange Commission.
Other useful information may include mortgages or liens against
each title-holding company or individual. This task differs from
the financial assessment task which will be discussed briefly later,
in that the financial status complies existing data while the finan-
cial assessment is an analysis of the existing data.
Report Preparation
   Once these basic tasks have been completed to determine the
history of the site and to  formulate a list of PRPs, a comprehen-
sive written summary needs to be prepared. The report should
discuss the research performed,  research results, data gaps and
recommendations for additional research to fill the data gaps. The
report also needs to include all sources contacted, even if infor-
mation was not available, to eliminate repeating the effort at a later
date. Sources which are identified as possibly having information
but are not contracted for some reason, also need to be identified.
The reports must be fully documented, attributing all information
presented to specific sources.

SPECIALTY TASKS
   The remaining 18 tasks discussed in the PRP manual may be
useful in some searches,  but not in others. In almost every case,
at least one of these tasks is used to further characterize the rela-
tionship  between PRPs and the site. If the researcher knows at the
beginning of the search that some of the specialty tasks may be
useful in identifying  PRPs, those tasks should be completed
concurrently with the  10 basic tasks. In other situations, it  may
be useful to complete the 10 basic tasks, review the findings and
determine whether  additional research is necessary.
   The manual categorizes  the 18  specialty tasks into four  sub-
sections: (1) obtaining specialized information.  (2) performing
waste  stream  comparisons,  (3)  creating  databases  and  (4)
performing other tasks.

Obtaining Specialized Information
   The tasks in the first subsection, obtaining specialized informa-
tion, includes eight tasks that may provide more information about
a site and its PRPs. These tasks  are:

   Aerial photographs
   CERCLA subpoena authority
   Field survey
   PRP file review
   Private citizen/PRP interviews
   Private investigations
   Site enforcement tracking systems
   Site sampling

Aerial Photographs
   Aerial photographs can be used to determine the relationship
between  a site and its surroundings. Taken over a period of time,
aerial photographs can help characterize the chronological develop-
ment of  a site. These photographs can help determine if surface
impoundments or other disposal areas were active during a PRP's
tenure as owner or operator.

CERCLA Subpoena Authority
  The CERCLA subpoena authority task is based on the adminis-
trative subpoena provision of SARA, which gives the U.S. EPA
the power to require the attendance and testimony of witnesses
and the production of documents. An administrative subpoena is
most useful in two situations:  (1) where preliminary information
has already been gathered and the researcher wishes to question
a particular person in detail and (2) where expedited enforcement
is being considered.

Field  Survey
  The next task, field survey, can be used to gather additional
information through visiting the site  and  its surroundings.  This
task is generally a routine part of site investigations and remedial
investigations. PRP leads may come from many sorts of observa-
tions such as abandoned vehicle license plate numbers, drum labels
or types of neighboring facilities.

PRP File Review
  The PRP file review task is used to: (1) locate and obtain relevant
documents in the possession of PRPs  and  (2) gather information
about the site's history and other PRPs. Relevant documents may
include customer lists, gate logs, customer correspondence, invoices
or operating logs.
Private  Citizen/PRP Interviews
  Interviewing private citizens and PRPs may be an effective means
of gathering information on site operations and history, PRPs,
other sources of information and previously identified data gaps.
Oftentimes these persons have first-hand knowledge of site activities
that occurred many years earlier or activities that are not document-
ed in  site records.
Private Investigations
  Private investigators (PI) generally are hired to locate PRPs or
obtain financial information about PRPs. Pis are used  in poten-
tially dangerous situations or if the research can be conducted more
efficiently or economically because the PI is located in or familiar
with the local community.

Site Enforcement Tracking Systems
  The next task involves using Site Enforcement Tracking Systems
(SEIS), a database maintained by the U.S. EPA's Office of Waste
Programs Enforcement that indicates whether a party has been sent
a notice letter. The database information may include the PRP's
name and address, site information and U.S. EPA site contacts.
From the SETS database, the researcher can determine if a PRP
is involved with other hazardous waste sites.

Site Sampling
  Site sampling is conducted to connect a specific, identified waste
type with a  PRP through chemical analysis of samples collected
at the site. Rarely is site sampling directly a part of the PRP search;
however, the researcher may have input on future sampling con-
ducted during the site investigation or  remedial investigation if
previous sampling is not sufficient to support the researcher's needs.

 Waste Stream Comparisons
  The second category of specialty tasks is performing waste stream
comparisons. This category includes  three tasks:

• Industrial safety
• Waste stream inventory
• Process chemistry analysis

Industrial Survey
  The industrial survey task is used to determine those parties who
may have contributed to the  site contamination by determining
which industries in the area generate, transport or dispose of
hazardous substances or wastes. The purpose of the survey is site-
specific. For sites such as landfills, the purpose may be to identify
                                                                             EPA POLICY PAPERS AND GUIDELINES     23

-------
parties who could have used the site. For sites with groundwater
contamination, the purpose may be to identify upgradient indus-
tries. This task is particularly useful when little information is avail-
able from documents, interviews and other usual sources or when
the site is in an area where neighboring facilities may have con-
tributed to the contamination.

Waste Stream Inventory
  The waste stream inventory task is used to compile an accurate
inventory of wastes that were stored or disposed of at a site by
reviewing all document such as waste stream records and operating
logbooks.  Knowing the types of contamination at  the site  is
necessary to identify the relationship between the site and PRPs.

Process Chemistry Analysis
  After identifying area industries through the industrial survey
task and determining the types of waste at the site through site
sampling and the waste stream inventory task, the researcher con-
ducts  the process chemistry analysis task. This task is used to
determine the types of wastes generated by the identified indus-
tries; these wastes are subsequently compared to the wastes at the
site. Once the researcher establishes a link between an industry and
wastes at the site, additional data gathering efforts can be initiated
to further identify an industry's specific waste handling activities.

Creating Databases
  The third specialty task  category, creating databases, includes
three tasks:

• Correspondence Tracking Databases
• Inventory Databases
• Transactional Database

  These  three databases are those most commonly used in PRP
searches.

Correspondence Tracking  Databases
  Correspondence tracking databases are used to track the mailing
of and responses to notice letters and information requests.  The
types of information tracked can include the names of parties to
whom letters were sent, whether the letter was received, whether
the party responded and a brief summary  of  the response.

Inventory Databases
  The second type of  database,  the  inventory database,  is a
management system for organizing and controlling case document
files and for summarizing case-specific information  contained in
the documents. Examples of the services these systems can provide
include:

• An effective way of locating and retrieving documents by key-
word, subject, author or date
• A document control system to assure that documents are not
  lost or misplaced
• A means for assuring an orderly and timely  response to orders
  or exchanges
CONCLUSION
  In summary, the manual outlines 28 tasks that can be completed
during a  PRP search. Ten  of the tasks generally are performed
for all PRP searches, while the other 18 are  specialized tasks that
may be useful in some searches but not in others. Before any tasks
are started during a specific PRP search, the researcher must clearly
understand the objective of the search. The researcher must be
aware of: (1) the type of case, such as a landfill with many PRPs
(2)  the types of PRPs  possibly involved with the site (owners,
operators, generators, transporters); (3) specific needs to support
case development, such  as  identifying only corporate PRPs; (4)
information already available to  avoid  repetition  of previous
efforts; and (5) anticipated action at the site. This understanding
will assist the researcher in identifying those tasks necessary for
inclusion in the PRP search to obtain the require information in
an effective and efficient manner.

Transactional Databases
  The third type of database, the transactional database, is used
to store information contained  in transactional documents. The
nature of the waste disposal industry often requires that several
transactions be made before final disposal of a hazardous waste.
Summaries of the database  information can display, where avail-
able from documentation,  evidentiary information on the types
of wastes disposed of, the haulers and generators, the total volume
of each waste type, and the quantity of each waste type by genera-
tor or hauler.

Other Tasks
  The fourth category of specialty tasks is miscellaneous tasks not
directly related to PRP identification but often included in PRP
searches. These tasks are:

• Compliance history
• Financial  assessment
• Generator ranking
• Property appraisal

Compliance History
  The compliance history  task involves reviewing records and
information to identify violations of hazardous waste and other
environmental laws and regulations. A profile of the PRP's com-
pliance history can then be prepared.  Documents useful  for this
identification include inspection reports, violation notices, legal
actions and  correspondence with  regulatory agencies.

Financial Assessment
  Financial assessment results are used to project a PRP's capacity
to address an environmental problem or a violator's ability to pay
a penalty.  Knowing such  information, U.S.  EPA  can better
formulate an appropriate negotiation and litigation strategy. Both
the U.S. EPA Supcrfund Settlement Policy and the U.S. EPA Gvfl
Penalty Policy contain provisions regarding the ability to pay as
an enforcement criteria. The financial assessment goes beyond the
final status task, one of the 10 basic tasks, by including an analysis
of the information collected.

Generator Ranking
  The generator ranking task is used to rank generators by the type
and amount of waste disposed of at a site. This is an important
element of the U.S.  EPA's Superfund Settlement Policy which
provides that the quantity and type of wastes contributed to the
site by various PRPs can be considered in evaluating settlement
offers. Moreover, the U.S. EPA  has  committed to releasing
information about the volume and nature of wastes to PRPs to
facilitate settlement discussions. However, the accuracy of this
ranking system depends on  the completeness of the  records
available. The U.S. EPA can release information on the  volume
and nature of wastes only to the extent identified as being sent to
the site.

Property Appraisal
  The property appraisal task can be used to assess the monetary
value of certain contaminated real property to support remedial
actions evaluated  or undertaken in accordance with the National
Contingency Plan. Contemplated  remedial actions might  include
Fund-sponsored cleanup, possibly including purchasing land and
relocating residents. Appraisals of the property "as is"  (before
remedial action) and "as modified" (after remedial action) often
are required. Property appraisals can be included during a PRP
search if the researcher is  trying  to determine the assets of an
identified PRP.
24    EPA POLICY PAPERS AND GUIDELINES

-------
                Superfund  Innovative  Technology Evaluation  (Site)
                                            After  the First  Year

                                                 Ronald D. Hill,  P.E.
                                      U.S. Environmental Protection Agency
                               Hazardous Waste Engineering Research  Laboratory
                                                    Cincinnati,  Ohio
ABSTRACT
  The U.S. EPA's Office of Research and Development (ORD),
joining with the Office of Solid Waste and Emergency Response
(OSWER),  has initiated the SITE program. The SITE program
will help the U.S. EPA find, test and encourage the use of new
ways to destroy, stabilize or otherwise treat hazardous wastes rather
than just bury them in the ground. The overall goal of the SITE
program is to maximize the use of alternatives to land disposal and
containment at Superfund sites. To accomplish this goal, the
program will provide reliable cost and performance information
on technologies that offer an alternative to land disposal.  This
information will be generated by conducting pilot-scale or full-scale
demonstrations of alternative technologies at Superfund sites. The
SITE program has been functioning over a year.  Twelve tech-
nologies have been selected for the first series of demonstrations,
and 12 more are under review for the second series.

INTRODUCTION
  The 1986 amendments to Superfund1 create a comprehensive
program of research, development, demonstration and training to
promote the development of alternative and innovative treatment
technologies. The terms "alternative or innovative treatment tech-
nologies" are defined in the law to mean those technologies that
permanently alter the composition of hazardous wastes  through
chemical, biological or physical means so as to significantly reduce
the toxicity, mobility or volume. The terms also include technolo-
gies that characterize the extent and nature of site contamination
and technologies that assess the stresses imposed by contaminants
on complex ecosystems at a site.
  An Office of Technology Demonstration has been established
to carry out this research and demonstration program. The U.S.
EPA is authorized to enter into contracts, grants and cooperative
agreements  with public entities and private parties to carry out this
program. Project selection must follow a schedule and procedures
as specified in the law. The Agency can arrange for the use of actual
Superfund  sites  for  the research, testing and demonstration
projects. At least 10 field demonstration projects must be initiated
each year.
  In order to carry out this legislative mandate, the U.S. EPA has
established  the Superfund  Innovative  Technology Evaluation
(SITE) program. A strategy and program plan has been developed
and published.2 The SITE program was discussed at this Con-
ference in  1986.3
  The overall goal of the SITE program is to maximize the use
of alternatives to land disposal in cleaning up Superfund sites and
to encourage the development  and demonstration of new, inno-
vative treatment and monitoring technologies. The SITE program
has been designed to accomplish the following objectives which
correspond to the program's four parts:

• To identify and remove impediments to the development and
  commercial use of alternative treatment technologies
• To demonstrate at full scale the more promising innovative tech-
  nologies to establish reliable performance and cost information
• To develop procedures and policies that encourage selection of
  alternative treatment technologies for Superfund site cleanup
• To-accelerate and promote further development of promising
  innovative technologies that are not yet ready for full-scale
  demonstration

  This paper will concentrate on the demonstration aspect of the
program.

DEMONSTRATION PROGRAM
  The purpose of the demonstration and evaluation of selected
technologies is to develop performance, cost-effectiveness and relia-
bility data so that Superfund decision-makers can make sound
judgments as to the applicability of the technology to a particular
site. The results of the demonstrations should identify the limita-
tions of the technology, the wastes and media to which it can be
applied, the operating procedures and the approximate capital and
operating costs. The demonstrations will be carried out at full-scale
or at a scale that allows valid comparison and direct scale-up to
commercial size units. The duration of the demonstration will be
determined on a case by case basis but must be of a sufficient time
to adequately characterize equipment reliability and operational
variabilities.
  SITE demonstrations usually will be conducted at actual uncon-
trolled hazardous waste sites including Superfund sites, state sites,
sites from other Federal agencies and  private party sites.  Occa-
sionally, demonstrations will be conducted at U.S. EPA operated
test and evaluation (T&E) facilities. These T&E facilities would
have adequate pollution control and safety equipment in place so
that technologies could be evaluated over a wide range of operation
without pollutant releases outside of the facility. The T&E facilities
could provide for faster testing of smaller and moderate size units
under safe  and controlled conditions at a lower cost. The tests
conducted in these facilities would be on wastes brought in from
an actual hazardous waste site.
  No Federal, state or local permits will be required for any demon-
stration carried out at actual Superfund sites. However, the demon-
stration will comply with all legally applicable or relevant and
appropriate standards (as these terms are defined in Superfund).
                                                                          EPA POLICY PAPERS AND GUIDELINES    25

-------
Permits will be required for demonstrations at all other sites.
  The funding of the demonstration will be cost-shared between
the U.S. EPA and the applicant. The technology developer usually
will be expected to pay the costs to erect and operate the equip-
ment on-site and dismantle and remove the equipment at the end
of the demonstration. The U.S. EPA will pay for the costs of
sampling  and analysis,  quality assurance and quality control,
evaluating data and preparing reports. The U.S. EPA also will help
the developer obtain any required permits. Normally,  there will
be no exchange of funds between the Agency and the applicant.
In a  few  instances where  the technology is unique,  unusually
promising and high in financial risk, the U.S. EPA wiU consider
bearing a greater portion of the total project cost if the developer
is unable to obtain financing elsewhere.
  At the completion of the demonstration, the U.S.  EPA  will
undertake activities to encourage wide-scale use of the successful
technologies in the Superfund cleanup program. The Agency will
evaluate the application of these technologies to other sites  and
to wastes other than those tested. A technology transfer and tech-
nical assistance program will transmit results to potential users
within and outside of the Agency.
  The U.S. EPA has issued two requests for projects  under the
demonstration program—one in  1986 and the second in January
1987. It is the intent of the program to have at least one solicita-
tion per year. The first solicitation in 1986 produced 21 responses
from which 10 technologies were selected for the program. In
addition,  other technologies were added upon presentation  and
review by the U.S. EPA. The first set of technologies selected  into
the SITE  program (SITE-001) can be summarized as follows:

• Thermal—5
• Extraction—2
• Stabilization/Solidification—2
• LBiological—1

  These technologies are discussed later in  this paper.
  The second  solicitation (SITE-002) in 1987 resulted in 29
responses. Of these, 12 have been conditionally acceptable and three
others have potential merit and the developers has been asked to
rewrite their proposals. The conditionally acceptable technologies
fall into the following categories:

  Thermal—2
  Extraction—1
  Stabilization/Solidification—4
  Biological—4
  Chemical—1

  Final selection for SITE-002 will occur in  the fall of 1987, and
the demonstration should  take place  in late 1988.
  The SITE-001 projects should enter the demonstration stages
in the fall of 1987 and the spring/summer of 1988. A description
of these technologies follows.

TECHNOLOGIES
  Electric Infrared Furnace
  This thermal  treatment  process developed by Shirco Infrared
Systems, Inc., uses a primary chamber where infrared energy is
provided by electrically powered silicon carbide rods.  Tempera-
tures in the primary chamber can reach 1850ฐF, and residence times
can range  from  10 to 90 min. Organic materials volatilized in the
primary chamber enter a secondary chamber where combustion
is completed at temperatures up to 2300 ฐF and 2 sec gas residence
time.  Exhaust gases then pass through a venturi scrubber spray
tower.
  Over 40 infrared systems installed throughout the world are to
incinerate municipal and industrial sludges and regenerate activated
carbon. A portable pilot unit, housed in  a 45-ft van  trailer and
capable of processing 30 to 100 Ibs/hr, has been built and tested
at several hazardous waste sites. The SITE demonstration will in-

 26    EPA POLICY PAPERS AND GUIDELINES
volve a unit that can process between 100 and 250 tons/day to treat
PCB-contaminated soils at a site in Florida and a small unit to
treat soils at a site in a U.S. EPA Region V. Both systems are mo-
bile, are mounted on wheels for highway transportation by truck
and can be installed  at a site in several weeks.
  The infrared furnace is designed to incinerate soil or sludge con-
taminated with organic hazardous constituents such as  PCBs, pesti-
cides and dioxins. At Times Beach, Missouri, the portable system
achieved a destruction   and  removal  efficiency  (DRE)  of
99.999996% on  soil  containing 227 ppb of 2,3,7,8 TCDD. At
Joplin, Missouri, the  portable infrared furnace  was used to decon-
taminate creosote sludges and achieved a DRE of 99.99999% on
pentachlorophenol and 99.99% on napthalene.

Circulating Bed Combusior
  This advanced fluidized bed  incinerator was developed by GA
Technologies and is owned by Ogden Environmental Services. High
velocity air suspends a bed of solid particles  in the  combustion
chamber. This creates a highly turbulent combustion zone with
temperatures in the range of 1450ฐF to 2000ฐF into which solid
or liquid waste feeds  can be introduced. Solidified materials have
a residence time of 30 min, while gases have a residence time of
2 sec. There is  no afterburner. The suspended  bed materials (and
waste solids) that leave the combustion chamber are  recovered in
a cyclone and recirculated through the furnace. The exhaust gases
pass through a convective gas  cooler and flue gas filter.
  Approximately 30 circulating bed combustors  operate world-wide
burning materials such as high sulfur coal, peat, wood, waste oils
and municipal waste.  A transportable pilot unit with soil feed rates
of 400  Ib/hr achieved a DRE of 99.9999% on  soils contaminated
with PCB's at concentrations up to 12,000 ppm. A permit to
destroy PCBs has been issued to  this unit under the  TSCA. Test
burns were conducted on  a mixture of organic  liquids and several
halogenated compounds.  DREs of at least 99.99% were achieved
for Freon 113,  carbon tetrachloride, ethylbenzene, and zylene. The
SITE demonstration  will  be conducted with a 2 million BTU/hr
unit  located in San Diego, California. Wastes from one or two
Superfund  sites in California will be used for the demonstration.

Electric Pyrolyzer
  This  system, proposed by Westinghouse Waste  Technology
Services Division, is  designed for thermal destruction of organic
hazardous  wastes without  combustion. The system  operates
through rapid transfer of energy to the waste material causing disso-
ciation  of  the organic  molecules  into  individual  atoms.  The
pyrolyzer is designed  to operate at temperatures up to  3250ฐF. Gas
residence time can be controlled  as necessary  to  destroy organic
materials. Exhaust gases are treated in a water-cooled cyclone, bag-
house filter and venturi scrubber.
  In addition  to destroying organics, the pyrolysis system also
reduces metals to their elemental state and fixes all inorganic oxides,
sulfides and halides as liquid silicates. Solids fall into a molten bath
which,  when cooled, forms a vitrified  material with  very low
leaching potential.
  A prototype system has been constructed which can  process from
5 to 20 tons/day of solid waste material with up to 10% organics
and 25% water. The pyrolysis system can be in the form of a mobile
unit; for example, a 100-ton/day unit could be trailer-mounted and
set up in a few days. A demonstration site has not yet been selected.

Solidification and Stabilization
  In this process developed by  Hazcon, Inc.,  liquid  wastes or
sludges are mixed with a cementitious material (such as fiyash or
cement kiln dust) and a proprietary reagent.  The mixture becomes
hardened within 10 to 15 min, and the hazardous constituents are
microencapsulated with significantly reduced  leaching potential.
Hazcon states  that the process is applicable to  hydrocarbon waste
from refining operations,  petrochemicals, residue from oil-based
paint, solvents and greases and inorganic wastes containing heavy
metals. It can solidify wastes over a wide range of pH (wastes with

-------
pH from 0.95 to 14 have been treated in bench-scale tests), and
there is no limit on moisture content.
  The operation uses very simple and standard blending and mixing
equipment available from machinery suppliers.  A mobile field
blending unit which can be set up at a site in a few hours is availa-
ble. The resulting product can be extruded into drums or forms
creating solid blocks or it can be placed in situ.
  This solidification process has been used on wastes from several
chemical companies. Under the SITE program, the U.S. EPA is
interested in demonstrating the applicability of this process to
organic sludges and soils contaminated with organic materials. A
Superfund  site in  Region  III is being  proposed for the
demonstration.

Vacuum Extraction
  Terra Vac,  Inc., has developed a process for  in situ vacuum
extraction of volatile contaminants from soils and groundwater.
Extraction wells are placed in the unsaturated soil zone above the
water table. A vacuum source is applied to the well, and volatile
organic compounds  are drawn  off. The gas  stream is passed
through an activated carbon absorption system to control air
emissions. The system can also be used to pump groundwater to
the surface for treatment by spray aeration.
  Vacuum extraction should provide advantages in situations where
excavation of contaminated soils is impractical or very costly. It
was used to clean up a spill of carbon tetrachloride from a ruptured
underground storage tank. The vacuum extraction rate of carbon
tetrachloride reached 250 Ib/day from  an unsaturated soil zone
300 ft deep. After 30 months of operation, more than 70% of the
spilled volume was removed from this zone. In other applications,
methylene chloride and gasoline were successfully extracted from
contaminated soils. A SITE demonstration is being developed in
Region I to evaluate this process on soils contaminated with volatile
organic compounds, hydrocarbons and  solvents.

Pyroplasma System
  The pyroplasma process pyrolyzes wastes using a thermal plasma
field. The heart of the system is a plasma arc torch which produces
a thermal plasma with temperatures of  more than 9000 ฐF.
  Waste liquids are injected directly into the plasma where the
molecules are broken into their atomic states and the atoms then
recombine to produce hydrogen,  carbon monoxide, nitrogen,
hydrogen  chloride, paniculate carbon, carbon dioxide, ethylene
and acetylene. The product gas is scrubbed with caustic  soda to
neutralize and remove  the acid gases and remove particulates.
  The pyroplasma system can destroy any pumpable liquid organic
waste. The system has been operated in a series of tests using methyl
ethyl ketone, methanol, ethanol, carbon tetrachloride and Askarel.
The unit is not designed to destroy solids, although it  can handle
up to 40% solids if they are pumpable  and  can pass through a
200-mesh  screen. Heavy metals pass through the system in the
scrubber water.
  A mobile unit mounted on a 48-ft trailer is available and can
process  2 to 3 gal/min.
  Two SITE projects will use this technology. In one project, the
New York State Department of Environmental Conservation and
Pyrolysis Systems, Inc., will use a mobile plasma arc unit for the
destruction of sludges from Love Canal. A  second project will
evaluate a larger unit developed by Westinghouse Plasma Systems.
A site for the demonstration has not yet been made.

Oxygen Enriched Burner
   A proprietary burner developed by Advanced Combustion Tech-
nologies, Inc., uses pure oxygen in combination with air and natural
gas to destroy liquid hazardous waste. The use of oxygen allows
higher temperatures which improves the kinetics of waste destruc-
tion, allows higher waste throughput and reduces stack gas volume.
Solids and sludges can be coincinerated when the burners are used
in conjunction with rotary kiln furnaces.
  The burner is capable of achieving temperatures as high as
5000 ฐF and should effectively destroy wastes which require high
temperatures  for destruction (e.g., halogenated organics) or have
low heating values (small amounts of organic constituents in an
inert base). Liquid waste particle size must be less than 1/32  in.,
and blending may be required to reduce viscosity.
  A SITE demonstration of this burner is being planned  at the
U.S. EPA Combustion Research Facility in Jefferson, Arkansas.

In Situ Stabilization
  A SITE demonstration is being planned to evaluate the in  situ
stabilization and solidification of soils contaminated with PCBs.
The  technology employed will  be  a chemical fixation process
developed by International  Waste Technologies which bonds
organic and inorganic compounds in "macro molecules" which
are resistant to acids and naturally-existing deteriorating factors.
After a test program to characterize the stabilized products, spe-
cial drilling equipment developed in Japan will be used to inject
and blend the stabilization medium into the contaminated soils.
Permeability,  density and leaching tests will be conducted. A PCB,
contaminated soil site in Florida is being proposed for the demon-
stration.

Biological Degradation of Organic  Contaminants
  Biodegradation of organic contaminants in sludges and soils
using proprietary, naturally-occurring microorganisms developed
by Detox Industries, Inc., will be evaluated in another SITE demon-
stration. Detox has adapted microorganisms to biodegrade PCBs,
PCPs, creosote, oils, phenolics and PAHs. Particular naturally,
occurring microorganisms are cultivated for accelerated growth  and
applied to  the  contaminant which is  metabolized  into carbon
dioxide, water, and cell protoplasma.
  A test  project was reviewed by U.S. EPA's Region VI where
500 Ib. of PCB, contaminated sludges, soils and water were treated.
PCB levels were reduced to 4 ppm and as a result, the process  was
approved for use for the destruction of PCBs in the Region VI
states.
  A SITE demonstration on PCB-contaminated soils and sludges
and PHA, contaminated material from Superfund sites is being
planned.

REFERENCES
1. Superfund  Amendments and Reauthorization Act of 1986, U.S.
   Congress.
2. U.S. EPA, Superfund Innovative Technology Evaluation (SITE) Strategy
   and Program Plan, EPA 540/G-86/001, Washington, DC, Dec. 1986.
3. Hill, R. D., White, D.C.  and Ogg, R. N., "Superfund Innovative
   Technology  Evaluation Program," Proc., Seventh National Conference
   on the Management of Uncontrolled Hazardous Waste Sites, Washing-
   ton, DC,  Nov. 1986. 356-360.
                                                                             EPA POLICY PAPERS AND GUIDELINES    27

-------
          A  Compendium  of  Superfund  Field  Operations Methods
                                                  James  B. Moore
                                                 CH2M HILL,  Inc.
                                                  Reston, Virginia
                                                Robert E. Stecik, Jr.
                                                  NUS Corporation
                                                 Arlington, Virginia
                                                      Lisa Feldt
                                     U.S.  Environmental  Protection  Agency
                                         Hazardous Site Control Division
                                                  Washington, D.C.
 ABSTRACT
   A recent guidance document entitled A Compendium of Super-
fund Field  Operations Methods,  or  OSWER  Directive  No.
 9355.0-14, has been prepared for the U.S. EPA. This document
 provides a significantly reduced compilation of the many field
 operations methods that are being implemented during remedial
 response activities at hazardous waste sites. The techniques and
 methods described within the document are those typically
 associated with the various field  activities as performed during
 hazardous waste field investigations. The detailed text provided
 on the field methodologies is supplemented by separate discussions
 of the documentation and support activity functions also relevant
 to the successful performance of these remedial field activities. This
 additional information pertains to project planning and manage-
 ment, health and safety, quality control, laboratory interface,
decontamination and data validation.
   The compendium was written primarily to assist the U.S. EPA
Remedial Project Manager and Contractor Site Manager. The ob-
jective of the compendium is to provide these individuals with a
summary of  field techniques and documentation procedures that
can be easily referenced during project planning activities. The com-
pendium is envisioned as a reference manual used to reduce the
regeneration of procedures within Quality Assurance Project Plans
and Sampling Plans as well as to ensure a consistent  approach.
This, too, directly  relates  to  streamlining the  project planning
phase. The procedures  within the document are not intended to
serve as stand-alone standard operating procedures, but as a
method which may be applicable to the project  once site-specific
modifications are made.

INTRODUCTION
  In 1984, the U.S. EPA decided  to undertake  an assignment to
summarize useful field and documentation methods typically
employed during the performance of hazardous  waste field inves-
tigations. Thus, A Compendium of Superfund  Field Operations
Methods was developed. Webster's Third International Dictionary,
Unabridged, defines compendium as "a brief compilation or com-
position consisting of a reduction and condensation of the subject
matter of a larger work" or "a work treating in  brief form the
important features of a whole field of knowledge or subject matter
category." While  the reader may  take exception to the word
"brief" to describe this  1,334-page compendium, the two volumes
provide only a capsule of the technical information frequently used
on a Superfund project and not necessarily the policy and adminis-
trative information typically needed.
  Also, the compendium is to be referenced and modified consis-

28    EPA POLICY PAPERS AND GUIDELINES
tent to the direction provided in the "Data Quality Objectives for
Remedial  Response Activities"  guidance (OSWER  Directive
9355.0-7B) and "Field Methodology Catalog" (under preparation).
It is hoped that the consolidation of these three documents will
provide the rationale for field and documentation method selec-
tion, the limitations and uses of the method and finally guidance
to the appropriate site, specific modifications of the method. The
primary users of the information are expected to be the U.S. EPA
Remedial Project Managers (REM) and Contractor Site Managers.
These individuals are the nucleus of the  project planning phase
and have responsibility for successful day-to-day work execution.
  The compendium has  undergone the scrutiny of a three-tiered
review process. In early 1985, the preliminary draft was completed
and submitted to a workgroup review made up of U.S. EPA Head-
quarters and selected Regional personnel. As a result of the work-
shop review, the text was expanded to provide a greater level of
detail of the material discussed. It also was updated with recent
developments and innovations. Soon afterwards, a draft version
of the compendium was  submitted for a full U.S. EPA  Regional
review. Also, the U.S. EPA REM Contractors provided assistance
in reviewing the draft document. This review provided significant
information regarding the specific  modifications to the procedures
necessary at the Regional level. These comments  were easily in-
corporated within the document  under  "Region-Specific Vari-
ances."  The final review conducted  prior to  printing was an
interagency review. Only a relatively small number of comments
resulted  and were immediately addressed.

USE OF THE COMPENDIUM
  Each section of the compendium is developed around similar
type subsections including:

  LScope and  Purpose
  Definitions
  Applicability
  Responsibilities
  Procedures
  Region-Specific Variances
  Information Sources

  The "Scope and Purpose" section deals with a summation of
the information presented under the "Procedures" and  discusses
how they can be applied within the  project. The "Definitions'
section identifies the frequently used terms and acronyms used
within the section. The "Applicability" section describes how the
method  is  to be integrated  within the project. The "Responsi-
bilities" chapter  lists individuals  by type and their associated

-------
 responsibilities for implementation of the method. "Procedures"
 details the operation of the method. The "Region-Specific Vari-
 ances" section identifies Regional preferences and modifications
 to the procedures discussed. And, finally "Information Sources"
 lists the sources consulted to develop the information within the
 text.
   The information  contained  within the compendium provides
 coverage not only of field activities, but also of documentation
 and support activities. Table 1 provides a breakdown of the docu-
 ment, by section, as relevant to either of these activities. Both types
 of activities must be addressed during the preparation of project
 planning documents. The compendium can be used as a reference
 during the planning process when it is appropriately supplemented
 by site, specific modifications. For example, a quality assurance
 project plan (QAPP) could present techniques for gathering data
 on chemical concentrations in fish tissue in a manner shown in
 Figure 1. The information often provided within a QAPP refers
 to Region-specific variations. The compendium does its best to
 identify these variances; however, users of this compendium are
 encouraged to consult with the appropriate U.S. EPA official to
 obtain the  most current variations to the methods listed.
   As a final subsection to each topic, the compendium indicates
 a variety of information sources utilized during the preparation
 of the particular section. Most of the sources are typically availa-
 ble at city or university libraries. These sources also will be made
 available at the U.S. EPA library.

                            Table 1
             A Compendium of Field Operations Methods
     Field Activity  Section
 Implementing Field Activities


 Field Methods for Rapid
 Screening for Hazardous
 Materials

 Earth Sciences

 Surface Water Hydrology

 Meteorology and Air Quality

 Biology/Ecology

 Land Surveying, Aerial
 Photography,  and Mapping

 field Instrumentation
   Documentation and Support
        Activity  Section
Preparation of Project
Description and Statement of
Objectives

Sample Control, Including Chain
of Custody
Laboratory Interface

Sample Containers  and
Preservation

Earth Sciences Laboratory
Procedures

Data Reduction,  Validation,
Reporting, Review, and Use

Document Control
                                Corrective  Action
                                Quality Assurance Audit
                                Procedures
                                Quality Assurance Reporting
Field Activity and Documentation and Support Activity Breakdown

 FIELD ACTIVITIES
   Relevant discussions pertaining to field activities are addressed
 within the following nine sections of the compendium:

 •  Implementing Field Activities
 •  Field Methods for Rapid Screening for Hazardous Material
 •  Earth Sciences
 •  Surface Water Hydrology
 •  Meteorology and  Air Quality
 •  Biology/Ecology
 •  Specialized Sampling Techniques
 •  Land Surveying, Aerial Photography and Mapping
 •  Field Instrumentation

   A brief synopsis of each section, including a general descrip-
                                    TASK 1. BIOTA EVALUATIONS

                                    Subtask  l.A. Electrof iahina in Manq Creek
  A. Limitations  and Application—Subsection  12.6.3.3,  Aquatic
     (Freshwater)  Field  Methods Summary, pp. 12-24 and  12-25,
     Section 12,  Revision No. 0,  Compedium  of Field Operations
     Methods (COFOM fO)
  B. Sampling    Techniques—Subsection   D2,    Electrofishing,
     Appendix 12A, pp. 12A-32  through 12A-35,  COFOM fO.
     Modifications.   Only  carp will  be  collected.    Specimens
     smaller than 8 inches in  length and  2  pounds  in  weight
     will be released.  Any specimens  caught  below 14th  Street
     Bridge will be released.   See site safety plan for boating
     and collection safety  procedures.
  C. Laboratory Techniques—Subsection 12.6.3.3, pp.  12-23 and
     12-24,  Subsection E4,  Appendix 12A,   pp.  12A-41  through
     12A-46, COFOM |0.
     Modifications.  See CLP SAS in Task  4,  Sample Analysis.

                           Figure 1
                  Quality Assurance Project Plan
                        Example Citation

tion and a summary of the information contained within the
section, is provided below.

Implementing  Field Activities
  The information discussed  within  the  "Implementing Field
Activities" section addresses several areas involved in the implemen-
tation of fieldwork. This section provides general information on
the following topics  in individual subsections that identify  their
scope and purpose, definitions and applicability.

The Control of Contaminated Materials
Generated During Fieldwork
  General reference information is provided on the proper manage-
ment, storage  and disposal of contaminated materials. The dis-
cussion includes  sources of contaminated materials (including
decontamination solutions, disposable equipment, drilling muds
and well-development fluids, and spill- contaminated materials)
and containment methods.

Organization of the Field Team
  The team components and duties of team members are described
and guidelines are provided for the numbers of members  that are
necessary for the field team to  safely meet the stated goals of the
investigation.  In  normal fieldwork, eight roles are commonly
required for a field investigation team: Site Manager, field team
leader, site safety officer, personnel decontamination station oper-
ator/equipment specialist, communications supervisor, initial entry
party, work party and emergency response team.

Decontamination
  A general discussion of decontamination  issues  is provided,
including a list of detailed guidance documents for decontamina-
tion methods,  techniques, procedures, equipment and solutions.

General Health and Safety Considerations
  Several standard Health &  Safety  field procedures that are
normally used in the  conduct of remedial  response activities
(including site  safety plan, general safety practices and site survey
and reconnaissance) are outlined, and the OSHA H&S requirements
for each procedure are described.

Field Methods for Rapid Screening for
Hazardous Material
  This section  provides an overview of current techniques used by
contractors to rapidly screen the hazardous waste material at waste
sites. Field analytical techniques are used whenever the data quality
objectives specify Level I and  II analytical support as adequate.
Also described are the functions  and capabilities  of available
analytical instrumentation, and some  suggestions are  made
regarding analytical protocols for mobile laboratories. Several rapid
                                                                              EPA POLICY PAPERS AND GUIDELINES    29

-------
screening procedures for inorganic and organic compounds are
described.  Several approaches  to determining inorganic  com-
pounds, including field test kits, are discussed as well as three broad
categories of equipment used for field analysis and screening of
organic compounds: (1) portable, total organic vapor monitors;
(2) portable, selective organic instruments; and (3) mobile, selec-
tive organic instruments.
   The procedures  discussed in  this section have been  used for
several purposes including screening the site to determine the level
of safety  required for personnel  working at  the site; screening
samples to determine which compounds, or groups of compounds,
should be specified for further analysis, usually under the CLP;
and  screening  for characterizing material for removal and  in
refining the sampling plan to more precisely determine the number
and type of samples to be taken. By using field screening, changes
in sampling can occur while the field team is mobilized, rather than
waiting for data to return from CLP analysis. Field screening tech-
nique, such as the removal of drums, lagoons, pits, ponds  and other
waste sources, allows testing for compatibility and disposal category
classification before disposal.
   Attached to  this section is an  appendix entitled "Protocols,
Reporting and Deliverables." Its purpose is to address several
methodologies  that have been used  in  screening samples  at
hazardous waste sites. General guidelines for sample preservation
are presented,  and the User's  Guide to the CLP is cited as  a
reference information source for U.S. EPA protocols for preserving
inorganic samples. The appendix provides listings for: (1) mobile
laboratory protocols for organic analyses; (2)  TCL estimated
detection limits; (3) inorganic analysis protocols; and (4)  operating
procedure for XRF analysis  of soils and tailings with the Colum-
bia X-Met 840 Analyzer.
   In July 1987, the CLP will publish a Field Methodology Cata-
log that will contain detailed discussions of field analytical methods.
The CLP  catalog will provide a consolidated reference for use by
U.S. EPA, contractors, state and local agencies and PRP who will
be conducting field analysis.

Earth  Sciences
   Because this section is so  large, it  is organized into five main
topics:

   Geologic Drilling
   Test Pits and Excavations
   Geological Reconnaissance and Geological Logging
   Geophysics
   Groundwater Monitoring

Geologic Drilling
   Although  this discussion  focuses  on  drilling  for  sampling
purposes,  it is important to recognize that borings are also required
for in situ testing of subsurface materials and groundwater and
to allow installation of monitoring devices  including wells. Guide-
lines are provided for the selection of the most appropriate method
or combination of methods for multipurpose borings. The general
applicability  of specific optimization techniques is discussed, as
well as routine soil drilling and sampling techniques.
   A descriptive list is  provided  that details  five general steps
required for  the planning, selection and  implementation of any
drilling program. The steps involve: (1) review of existing drilling,
related, site-specific information; (2) development of a site, specific
health and safety program; (3) preparation of the drilling plan and
contract; (4)  field implementation and  decontamination; and (5)
reporting. General considerations for the selection and implemen-
tation of soil  drilling  and  sampling  methods  are  also  listed:
(1) prevention of contaminant spread; (2) maintenance of sample
integrity; (3) minimization of disruption of existing conditions; and
(4) minimization of long-term impacts.

Test Pits and Excavations
  Reference materials and general guidelines are provided for con-
                                                            ducting test pit and trench excavations: (1) test pit and trench
                                                            construction; (2) sampling techniques; (3) backfilling of trenches
                                                            and test  pits;  and (4)  decontamination.  Additional  specific
                                                            information is described regarding routine test pit or trench exca-
                                                            vation techniques, including health and safety concerns, machine-
                                                            dug excavations and the handling of hazardous materials brought
                                                            to the surface by excavation equipment.

                                                            Geological Reconnaissance and Geological Logging
                                                               A description  of the basic methods, procedures and activities
                                                            to be accomplished or  considered for a geological reconnaissance
                                                            are summarized. Also, industry standards for geological logging
                                                            of soil or rock materials derived from subsurface investigations,
                                                            as well as the applicability of geological reconnaissance and of geo-
                                                            logical logging, are addressed.

                                                            Groundwater Monitoring
                                                               General information is provided on equipment and materials,
                                                            as well as procedures for water wells, lysimeters, piezometers and
                                                            tensiometers,  groundwater sampling  equipment,  water, level
                                                            measurement  devices,  field parameter measurements, filtration
                                                            materials for  well construction and groundwater sampling con-
                                                            siderations.

                                                            Geophysics
                                                               General guidance is  provided  for the planning, selection and
                                                            implementation of geophysical surveys that may be conducted. Six
                                                            commonly used  methods are discussed from the standpoint of
                                                            applicability to site investigation, procedures for implementation,
                                                            survey design, and miscellaneous method-specific considerations.
                                                            The methods discussed include electromagnetics, electrical resis-
                                                            tivity, seismics, magnetics, ground penetrating radar and borehole
                                                            geophysics.
                                                               An introduction to  the basic borehole geophysical techniques
                                                            is presented also. The general logging categories discussed are elec-
                                                            trical, nuclear, sonic and  mechanical. A very basic description of
                                                            the log, the parameters that affect response and the sensing devices
                                                            are presented to aid in evaluating the applicability  of  logging
                                                            functions.
                                                               Six appendices are provided as an attachment to the geophysics
                                                            section. They contain rigorous theoretical explanations of the six
                                                            respective methods that are commonly used for geophysical  surveys.
                                                            The appendices are intended for use as reference sources by those
                                                            readers who may desire  a  more thorough  explanation  of each
                                                            method.

                                                            Surface Hydrology
                                                               Two general topics are presented in this section, "Flow Measure-
                                                            ment" and "Sampling." General  guidance is provided for the
                                                            planning, method selection and implementation of surface flow
                                                            measurement  techniques. The discussion includes  treatment of
                                                            general methods and applications and alternative flow measure-
                                                            ment techniques. It also addresses the following issues which are
                                                            common to all surface flow measurements at or near hazardous
                                                            waste sites: (1) preventing the spread of contamination; (2) mini-
                                                            mizing the risk to health and safety; (3) maintaining a high level
                                                            of accuracy in measuring flows; (4) causing the least possible dis-
                                                            ruption to on-site activities; (5) reporting all readings in an or-
                                                            ganized   fashion as  required  by  the sampling plan; and
                                                            (6) reducing, where possible, any additional long- and short-term
                                                            impacts. The information that is provided regarding sampling tech-
                                                            niques serves as general guidance for the collection of surface water,
                                                            sediment samples and sludge.
                                                            Meteorology and Air  Quality
                                                               A description is provided regarding the meteorological data that
                                                            are  required  to make preliminary  (screening)  assessments  of
                                                            exposure to hazardous air  pollutants before site-specific monitoring
                                                            data are available. Similarly, the meteorological data requirements
                                                            for conducting analyses of more refined air quality modeling are
                                                            explained in terms of using both representative off-site and site
30
EPA POLICY PAPERS AND GUIDELINES

-------
specific data. The procedures  for  obtaining  the  appropriate
meteorological information from both existing sources and by con-
ducting site-specific monitoring programs are identified, and the
requirements for meteorological input data, screening model selec-
tion and meteorological data selection are discussed.
  Screening models are used to provide a conservative estimate
of the air quality impact of a specific source or source category.
The section discusses the different objectives that screening model
analyses can be used to determine, depending on the level of refine-
ment. Descriptions of procedures for meteorological data collec-
tion are provided, covering several techniques including visual
observations, representative sources such as NWSstations and in
situ sensors. Guidelines for monitor placement also are provided.
  A discussion on the general air and gas sampling  methods for
determining air quality is presented. It describes the methods and
equipment necessary for real-time air quality monitoring in the field
and for collecting air samples for  laboratory analysis.
Biology/Ecology
  The general types of field  and laboratory activities that can be
used to assess  biological  or ecological impacts resulting from
remedial response activities are described. This section comprises
four basic components: (1) introductory remarks regarding bio-
logical and ecological evaluations of hazardous waste sites; (2) a
summary of the methods and applications that have been used to
date, and their limitations; (3) a list of references to lead a user
to more details about methods; and (4) additional method details,
included as  an  appendix.
  This section discusses terrestrial, aquatic (freshwater) and near-
shore marine environments. A description is provided of the types
of biological field sampling techniques and laboratory  analyses that
have  been  used or are being used  in biological or ecological
assessments. Procedures mentioned  include techniques used to
determine the presence of toxic substances, general field collec-
tion techniques, specific field methods (for terrestrial fieldwork,
aquatic-freshwater fieldwork, marine fieldwork, vegetation, ter-
restrial  animals and aquatic vertebrates) and laboratory tests and
analyses.
  An appendix of collection and processing techniques includes
detailed discussions of collection techniques for vegetation, ter-
restrial  vertebrates,  aquatic macroinvertebrates and  fish. The
appendix also  contains directions  for  biological field sample
processing and preservation.

Specialized Sampling Techniques
  Wire samples are used to document the presence of carcinogenic
substances or other toxic materials. A discussion of the steps re-
quired to obtain a wipe sample is provided. Wipe sampling can
help to provide a picture of contaminants that exist on the surface
of drums, tanks, equipment or buildings on a hazardous waste site
or that  exist in the homes of a populace at risk. The discussion
on procedures includes a listing of the equipment required and a
seven-step method for obtaining a wipe sample.
  General guidance is provided for the planning, method selec-
tion and implementation of sampling activities used to determine
the potential for human exposure to contaminants that are present
in a residential environment. This discussion pertains to sampling
techniques that are similar in collection methodology to other types
of samples, such as environmental soil and water, but are biased
to emphasize potential human exposure to contaminants moving
into the residential environment. The sensitivity involved in human
habitation sampling is emphasized as well as the responsibility of
community relations, health and safety and sample collection per-
sonnel to be aware of the  resident's perspective. Several types of
samples are described that can be related to the potential for hu-
man exposure to  contaminants  in the residential environment.
These samples include: (1) vacuum bag; (2) air conditioner filter;
(3)  dust   sweep;  (4)  sump   or  drain  sediment;  and
(5)  lint  trap.
  Also, general  information  is  provided  on  performing
2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) analysis. However,
the procedures used in sampling TCDD are frequently updated;
the U.S. EPA Regional Sample Control Center (RSCC) should be
contacted about preferred collection techniques. Procedures for
TCDD sampling are described, including discussion of sampling
activities,  blending  procedure  and  field  quality  control
requirements.

Land Surveying, Aerial Photography  and  Mapping
  Information is provided for use in the planning and implemen-
tation of land surveying, aerial photography and mapping. The
methods of obtaining maps through field surveys, property surveys,
surveys of monitoring wells, aerial photography and photogram-
metric mapping are detailed. Also, the section discusses procedures
for general surveying  (including  third-order vertical survey,
property surveys, traverse computations and adjustments, level
circuit computations and adjustments and monitoring well surveys),
aerial photographing (including contracting for aerial photography
and photogrammetric mapping), remote sensing and hydrograph-
ic surveys.
Field  Instrumentation
  Basic information on operating various pieces of equipment that
are  typically used in the field is provided (i.e., calibration and main-
tenance). Field monitoring instruments are used whenever the data
quality objectives specify Level I and II screening analytical sup-
port as adequate. The objective of Level I analysis is to generate
data that are generally used to refine sampling plans and to estimate
the  extent of contamination at the site. This type of support pro-
vides  real-time data for health and safety purposes.
  This section provides operational descriptions of the Photovac
10A10,  the HNU PI-101, the  Organic Vapor Analyzer 128
(OVA-128), explosimeters, oxygen indicators, combined combus-
tible gas and oxygen alarms, vapor detection tubes, the Draeger
gas detector  model  21/31, radiation monitors and  personal
sampling pumps. Other monitoring devices discussed in this section
include electrochemical gas detectors, passive dosimeters and the
MINIRAM (Miniature Real-Time Aerosol Monitor). Detailed dis-
cussions of the field analytical methods described here can be found
in the Field Methodology Catalog.

DOCUMENTATION AND SUPPORT ACTIVITIES
  The remaining parts of this guidance document address infor-
mation pertaining to documentation and support activities as they
relate to the field activities. The information is contained within
the  following ten sections of the compendium:

  Preparation of Project Description and Statement of Objectives
  Sample Control, Including  Chain of Custody Laboratory In-
  terface
  Sample Containers and Preservation
  Earth Sciences Laboratory Procedures
  Data Reduction, Validation, Reporting, Review and Use
  Document Control
  Corrective Action
  Quality Assurance Audit Procedures
  Quality Assurance Reporting

  Provided below is a brief synopsis of each section, including a
general description and a summary of the information contained
within the section.

Preparation of Project Description and
Statement of Objectives
  Under remedial response activities, such as field investigations
and sample collection, a written work plan is required which should
include a project description detailing the objectives of the study.
Methods on the means of accomplishing this task are described
within this section. Information such as the collection and evalua-
tion of existing data is explained. Considerations  also are given
to the scheduling of activities including logic determination and
                                                                             EPA POLICY PAPERS AND GUIDELINES    31

-------
critical path.
  Supplementing the description of these activities, the text dis-
cusses the importance of outlining the intended data usage, iden-
tifying sample matrices and parameters and preparing a sample
design description and rationale. Later sections deal more specifi-
cally with the application of the Data Quality Objectives guidance
within the framework of the project planning process.

Sample Control, Including Chain of Custody
  Procedures for implementing sample control and chain  of
custody activities are described,  including  sample identification
tags,  sample traffic report  (TR),  chain-of-custody forms and
records, receipt-for-samples form, custody  seals, field notebooks
and corrections to documentation. The purpose of these proce-
dures is to maintain the quality of samples during collection, trans-
portation and  storage for analysis. More specific information
concerning sample control under the CLP is presented within the
User's Guide to the Contract Laboratory Program.
  Sample identification documents include: (1) sample identifica-
tion tags; (2) sample traffic reports; (3) chain-of-custody records;
(4) receipt-for-samples forms; (5) custody seals; and (6) field note-
books. The following additional forms are used for samples shipped
to CLP laboratories: (1) organic traffic reports; (2) inorganic traffic
reports; (3) high-hazard traffic reports; (4) SAS request forms; and
(5) dioxin shipment records (as applicable). Completed examples
of  these  forms and a descriptive list of the records that need to
be  maintained  for sample control activities are included.

Laboratory Interface
  This section is organized by CLP and non-contract laboratory
program to provide a clearer differentiation. There is no formal
non-contract laboratory program (non-CLP) run parallel to the
CLP. A non-contract laboratory is procured by a method other
than going through the SMO. The information provided on the
National CLP includes a summary of how to schedule analyses
through the CLP, the types of services provided by the CLP, the
paperwork involved in submitting samples to a CLP laboratory
and how to contact a CLP laboratory regarding final disposition
of analytical data. Also, guidelines are provided detailing how to
contact a  non-contract laboratory, the paperwork involved when
submitting samples to such a laboratory and  the resolution of ques-
tions once the analyses have been completed. A detailed discussion
of the entire CLP, including the CLP tracking system, can be found
in the User's Guide to the CLP.
  Guidelines indicate how to access and coordinate with the CLP.
General responsibilities  as related to laboratory interface are
described  for SMs, RPMs, RSCC, SMO and sampling personnel.
A list  of the documentation required for each sample collected is
provided, as well as a list of information sources and a list of CLP-
related definitions and abbreviations. A chronological description
presents the procedures used during a routine well sampling epi-
sode, including activities before sampling, sampling activities and
postsampling activities for both CLP and non-contract laboratories.
However, it is  important to  note  that several  region, specific
variances  are identified  regarding  these  sample coordination
activities, and each region should be consulted prior to undertaking
a sampling activity.

Sample Containers, Preservation and Shipping
  For greater clarity, this section is  presented as  two  topics:
(1) sample containers and preservation; (2) and packaging, labeling
and shipping. A chronological description  of the procedures for
a typical sampling episode is presented. The specific sampling
activity discussion involves: (1) representative sample collection;
(2)  low-concentration water sample preservation techniques; and
(3)  sample shipment for laboratory analysis. A description of the
procedures used to package, label and ship environmental and
hazardous samples is provided. The procedures described in this
section apply to samples collected at a waste site, and they must
be followed whether shipping to a  CLP laboratory or to a non-

32    EPA POLICY PAPERS AND GUIDELINES
contract laboratory.

Earth Sciences Laboratory Procedures
  This section identifies the laboratory procedures used to deter-
mine the physical and chemical properties of soil materials. Pro-
cedures are given for volumetric, strength and transport relationship
tests and  for testing  chemical properties. General guidance is
provided for the planning and implementation of laboratory testing
earth science materials for hazardous waste projects. The discus-
sion includes both a broad overview of the types of routine labora-
tory techniques available for use and a brief, general description
of their purpose and applicability.  Reference  listings are also
presented  for specific testing techniques and standards.
  Procedures are described for analyzing soil samples collected
during field investigations. The results of these procedures can be
used in soils engineering determination, contaminant  migration
evaluation and design considerations.
  Most of the procedures described are derived from the 1984
American  Society of Testing and Materials (ASTM)  Book of
Standards. Users of the compendium should review the most recent
ASTM procedures for  changes. Procedures are  described  for
sample handling, physical tests (volumetric, strength and transport
relationships), chemical tests (mineralogy, cation exchange capacity
and distribution coefficient) and laboratory  records. Guidelines
are presented to evaluate the suitability of a laboratory to conduct
the procedures.

Data Reduction, Validation, Reporting,
Review and  Use
  The data validation procedures that are specific to the CLP are
summarized, including qualitative and quantitative investigations,
reduction, validation and data review and use. The section also
describes sources of data errors and  approaches to reduce these
errors.
  The  CLP  offers routine analytical services (RAS) that deliver
analyses of the Target Compound List (TCL) organic compounds.
Target Analyte List (TAL) inorganic parameters and dioxin. Special
analytical  services (SAS) also are available through the CLP. A
list of definitions and abbreviations for related terms is  provided.
  A systematic process to consider when measuring environmental
contaminants is recommended by the American Chemical Society
in ' 'Guidelines for Data Acquisition and Data Quality Evaluation
in Environmental Chemistry." This process considers the planning,
measurement,  calibration  standardization,  quality assurance,
statistical procedures and documentation needed for high-quality
analytical  chemical data.

Document Control
  Procedures are described for implementing document control
activities,  including project files, document identification and
numbering,  document distribution, revisions to documentation,
project logbooks, computer codes and documentation, corrections
to documentation, confidential information and disposition of
project documents. These procedures are designed to ensure that
documents are reviewed for adequacy, completeness, correctness
and release by authorized personnel. There also are procedures to
facilitate changes to documents; identification and distribution of
controlled documents; preparation, review, approval and issuing
of documents; document file; document number; and  inventory
documents.

Corrective Action
  A procedure  is described to implement and document correc-
tive actions.  Guidance relevant to several corrective action issues
is provided,  including limits for and control of data acceptability,
reviews, non-conformance, corrective action approval,  corrective
action  review and corrective actions  for data acceptability. The
corrective action program includes the investigation of the cause(s)
of significant or repetitious unsatisfactory conditions affecting the
quality of materials,  components or services,  or  the  failure to

-------
implement or adhere to required quality assurance practices.

Quality Assurance Audit Procedures
  The procedures for quality assurance audits are described. The
pre-audit meeting, audit performance, evaluation of audit findings,
post-audit meeting and audit reporting are addressed. Procedures
are described for audit schedules, qualification and certification
of quality assurance personnel, preparation for audits, conduct of
audits and audit follow-up.
  The following records are generated in support and completion
of the quality assurance audits for Superfund projects: (1) audit
schedules and revisions thereto, (2) audit qualification records, (3)
certification records (current and historical), (4) audit checklists
and audit guides,  (5) audit plan,  (6) audit reports,  (7) written
response to audit reports, (8) response evaluations and (9) records
of audit closure.
 Quality Assurance Reporting
   On a periodic basis, QA reports  are routinely issued to  the
 appropriate Site Manager and, as appropriate, to the responsible
 higher management. These reports summarize the QA/QC status
 of the projects and any conditions adverse to quality.
   Procedures for assessing the QA/QC status of project activities
 and the means of reporting such assessment are  described. These
 procedures  include assessment of measurement data accuracy,
 assessment of performance and systems audits, non-comformances
and assessment of quality assurance problems and solutions.

CONCLUSIONS
  The compendium contains a wealth of information pertaining
to both field activities and documentation and support activities.
This information typically is integrated into Superfund projects
during the project planning phase. The compendium is not intended
to serve as a stand-alone standard operating procedure, but to be
used as a reference source when appropriately supplemented by
other available guidance information, site-specific modifications
and regional variations.
  The compendium represents only a snapshot of methods  and
techniques that, in the rapidly evolving field of remedial response,
will undergo changes as new procedures are defined. Additionally,
methods that were not  included in this compendium because of
a lack of demonstrated  success at the time of writing may rapidly
emerge as methods of choice. The U.S. EPA's intent is to provide
annual updates presenting newly evolved methods and improve-
ments on old methods. As well, several of the procedures used on
Superfund  projects  are frequently updated and the  U.S.. EPA
should be contacted about preferred techniques.
  A regional workgroup is expected to be held annually to outline
upcoming revisions. Specifically, the first Revision (01) of the com-
pendium will contain the latest information published within the
CLP "Field Methodology Catalog." Pertinent comments, sugges-
tions and recommendations are strongly encouraged from the users
at any time.
                                                                              EPA POLICY PAPERS AND GUIDELINES    33

-------
                         Federal  and  State  Liability  Standards for
                         Superfund Response  Action  Contractors

                                                  Robert J. Mason
                                              Chief, Guidance Section
                                     U.S.  Environmental Protection Agency
                                     Office of Waste Programs Enforcement
                                                 Washington, D.C.
                                              Douglas W. Kohn,  J.D.
                                          Mark F. Johnson,  MBA,  ARM
                      Planning  Research  Corporation—Environmental Management,  Inc.
                                                    McLean,  Va.
 INTRODUCTION
   Subject to certain restrictions, Section  119  of the  SARA
 authorizes the U.S. EPA and other federal agencies to indemnify
 response action contractors' (RAC) for claims brought for negli-
 gent pollution  liability arising from response actions at sites on
 the NPL, or from removals conducted under SARA. RACs con-
 ducting removal and remedial response actions for the U.S. EPA,
 other federal agencies, states and PRPs at sites on the NPL,  are
 eligible for Section 119 indemnification.
   An important new provision of SARA is Section 119(a) which
 establishes a non-preemptive federal negligence2 liability standard
 for RACs participating in the  Superfund  program. Congress
 adopted a federal negligence standard for RACs as a partial solu-
 tion to liability concerns that the RAC community raised  during
 Superfund reauthorization.
   This paper briefly discusses the major issues  raised by RACs con-
 cerning liability during Superfund reauthorization and summarizes
 Congressional response to RAC liability concerns by outlining the
 major provisions  of Section 119 of SARA. The paper discusses
 the Congressional rationale behind the adoption of a federal negli-
 gence standard  for RACs based on a review of the legislative his-
 tory of Section  119 of SARA  and presents  the results of a
 preliminary U.S. EPA assessment of state RAC liability standards.
 Further, it discusses the current availability of pollution insurance
 for RACs working in the Superfund program. Finally, this paper
 discusses the enactment of specific statutory negligence liability
 standards for RACs in New Jersey, Louisiana and New York and
 the potential impact these state laws may have on availability of
 future commercial pollution liability insurance for RACs.

 LIABILITY  CONCERNS RAISED
 BY THE RAC  COMMUNITY DURING
 SUPERFUND  REAUTHORIZATION
  During  the  Superfund reauthorization  process  (hereafter
 reauthorization),  the RAC community raised several concerns
 which RACs contended impaired their ability  to adequately offset
 potential pollution liability arising from their response  action
 activities in the Superfund program. The RAC  community was con-
 cerned that potential pollution liability damage claims arising from
 the Superfund site response actions could subject RACs to poten-
 tially adverse liability in the future.  The major concerns raised by
 the RAC community during the Superfundreauthorization included
 the following

 Note: Any views or  opinions expressed  in this paper are the views of the
 authors  and do not  necessarily represent the views of the U.S. EPA.
• The potential application of strict3, joint and several liability4
  under federal law (for example, CERCLA) and some state laws
  to RACs participating in the Superfund cleanup program
• The inability of the commercial liability insurance market to pro-
  vide affordable and adequate pollution liability insurance cover-
  age to RACs involved in the Superfund cleanup program
• The inability of  the U.S. EPA to provide meaningful  indem-
  nification to RACs because of an absence of statutory authority
  to indemnify RACs
• The absence of a statutory source of funds to pay future RAC
  indemnification claims that may have violated the federal Anti-
  Deficiency Act

  Prior  to the passage of Section ;kx;l 19;ko;  of SARA, it was
uncertain whether RACs were subject to the strict, joint and several
liability  standard that generators, transporters and disposers of
hazardous waste are  subject  to under Section 106 and  107 of
CERCLA. At the outset of reauthorization, the RAC community
argued for limitations on RAC liability. They argued that liability
limitations were  necessary to protect RACs from future liability
claims stemming from CERCLA (or similar state laws) and com-
mon law. The RAC community proposed liability limitations in-
cluding:

• Statues of limitations
• Statues of repose
• Total exemption  from negligent liability, with RAC liability for
  gross negligence  and intentional  misconduct

  As reauthorization progressed, the RAC community argued that
RACs should, at a minimum, be held  to a negligence standard.
They argued that a negligence standard must be extended to all
the types of liability claims that RACs face, including federal, state
and common law. They also argued that without the adoption of
a uniform negligence  standard (across  all 50 states) and without
adequate risk transfer mechanisms (commercial  liability insurance
or federal indemnification), the following consequences might
result:

• Prudent RACs would not be willing to subject their limited cor-
  porate assets to  the uncertain legal environment surrounding
  Superfund response action work
• Availability of RAC capacity could  decline as prudent RACs
  leave the market
• The quality of Superfund cleanups could be reduced if quali-
  fied RACs refused  to perform Superfund response actions
34    LIABILITY

-------
U.S. EPA POSITIONS CONCERNING RAC LIABILITY
DURING SUPERFUND REAUTHORIZATION
  Throughout  reauthorization  the U.S.  EPA supported  the
adoption of a preemptive and uniform negligence standard and
discretionary indemnification for negligent pollution liability as
a comparable substitute for insurance for RACs. The U.S. EPA's
rationale for proposing these solutions to the RAC liability problem
included:

• Preemptive negligence—
  — A uniform  and preemptive negligence standard provides
     RACs with acceptable performance standards and provides
     the insurance industry with a traditional negligence standard
     that should provide incentives  to insure RACs prospectively.
• Discretionary indemnification—
  — Provides  an interim risk transfer mechanism to keep the
     Superfund program operative in the absence of adequate and
     affordable pollution liability insurance for RACs. Such an
     interim approach gives the Property and Casualty (P&C)
     insurance industry time to restore its financial capacity to
     underwrite pollution liability and develop sound underwriting
     approaches  for RACs participating in remedial responses.
  — Does not permanently replace the role of the P&C insurance
     industry.
  — Responds based on insurance market conditions.
  — Assures that adequate performance incentives for RACs are
     part of the  Superfund program.
  — Is consistent with administration and Congressional intent.
  The U.S. EPA  contended throughout reauthorization that one
of the primary causes of the current pollution insurance problem
faced by the RAC community and other industries is the cyclical
underwriting practices of insurers.  The U.S. EPA believes that,
during the late  1970s and early 1980s, high interest rates caused
many insurers to abandon traditional underwriting practices to rely
more on investment income from  high premium volume.  This
practice, known as cash flow underwriting, was widely employed
by U.S. insurers during this time period. During this period of high
interest rates in the U.S., many insurers employing cash flow under-
writing practices recorded record profits. However, as interest rates
declined, record underwriting losses experienced by P&C insurers
in 1984/85 caused many insurers to  reevaluate the profitability of
their insurance  lines. During this same time period, the U.S. and
international reinsurance markets withdrew from perceived high
risk liability lines, primarily because of their own record under-
writing losses.
  Beginning in 1985, due to limited insurance capacity from record
underwriting losses and a general lack of reinsurance support, many
U.S. P&C insurers withdrew from  many perceived high-risk lia-
bility insurance lines (for  example, pollution liability). When
insurance capacity is limited, insurers tend to withdraw from high
risk liability lines and save their limited liability insurance capaci-
ty for their more traditionally profitable insurance lines.
  Beginning in 1985, U.S.  insurers returned to sounder under-
writing and pricing practices and saved their limited  liability
insurance capacity for traditionally profitable insurance lines. The
practice of cash flow underwriting by insurers during periods of
high interest rates is not unique to the latest downturn of the current
P&C underwriting cycle. The U.S. EPA believes that this practice
has occurred repeatedly over the years in the U.S. In past P&C
underwriting cycles, after the cycle was complete and  insurers
restored financial capacity, insurers again began offering insur-
ance coverage for perceived high-risk liability lines like pollution
liability insurance.

SUMMARY OF SECTION 119 OF SARA
  Congress, concerned that prudent, qualified RACs may not par-
ticipate in the Superfund program without performance incentives
and adequate insurance or indemnification, considered several
options for addressing the concerns raised by the RAC community
during Superfund reauthorization. The major options considered
by Congress included the following:
• Modifying RAC liability standards
• Providing federal insurance and reinsurance
  Providing mandatory and discretionary federal indemnification
• Providing statues of limitation and statues of repose for RAC
  liability

Congress, after considerable debate on the available options,
incorporated Section 119 into SARA as a partial solution to the
RAC liability problems. Section 119 of SARA, as crafted by Con-
gress,  addresses many  of the  concerns raised  by  the  RAC
community  during Superfund  reauthorization. In  developing
Section 119, Congress specifically reviewed the U.S. EPA's existing
scope of RAC indemnification authority and RAC indemnification
authority and RAC liability standards. Congress responded by
incorporating into  SARA, Section 119, which establishes a non-
preemptive federal negligence  liability standard for Superfund
response action contractors and authorizes the U.S. EPA and other
federal agencies to indemnify RACs subject to certain requirements
and restrictions.  Section  119  includes  the  following  major
provisions:
• Exempts RCAs  from liability except  in  cases of negligence
  under all federal laws, but does not preempt state law.
• Provides the U.S. EPA and other federal agencies with discre-
  tionary authority to indemnify RACs  for claims  brought for
  negligent liability. The U.S. EPA and other federal agencies are
  not authorized to provide indemnification for strict liability
  where it exists at the state level.
• Authorizes indemnification of RACs working for the U.S. EPA,
  other federal agencies, a state under a contract or cooperative
  agreement, or any PRP.

• Section 119 can  be provided only—
  — To  RACs working in the Superfund program  and not to
     facilities regulated  under  the Resource Conservation and
     Recovery Act  (RCRA). This includes Publicly Owned Treat-
     ment Works (POTWs)
  — When a RAC  has made a diligent effort to obtain insurance
     and has found that it is unavailable
  — As a comparable supplement or substitute, to include deducti-
     bles and limits of indemnity for adequate insurance when
     such insurance  is  either  unavailable,  insufficient  or
     unreasonably  priced
  — As a comparable supplement or substitute, to include deduct-
     ibles and limits of indemnity, for adequate indemnification
     of RACs by PRPs, when such indemnification, as determined
     by the U.S. EPA, is either unavailable or insufficient
•  Section 119 indemnification payments will be made from the
  Hazardous Substance Response Trust Fund (the Fund). If suf-
  ficient funds are  unavailable  from the Fund or if  the Fund is
  repealed, authorization is provided to appropriate such amounts
  as may be necessary to make such payments. Amounts expended
  under Section 119 are considered governmental response costs
  for purposes of cost recovery.
•  Exempts Section  119 indemnification claim payments from the
  Anti-Deficiency Act.
*  Payment of a claim under Section 119 indemnification agree-
  ments RAC working for a PRP may be made only if the RAC
  exhausts all legal claims for indemnification against all PRPs.

CONGRESSIONAL RATIONALE BEHIND FEDERAL
NEGLIGENCE STANDARD FOR RACS IN SECTION 119
  Prior to the enactment of SARA, it was uncertain whether
CERCLA's strict, joint and several liability standard for owners,
generators or transporters could be applied to RACs in the event
of a release occurring during or following a response action. Under
                                                                                                           LIABILITY    35

-------
CERCLA's strict, joint and several liability standard, any RAC
working at a Superfund site was potentially liable for all removal
and remedial costs associated with the release or threatened release
of a hazardous substance from that site.3
  Congress explicitly removed  any doubt that  a  strict liability
standard could be applied to RACs under federal law in Section
(1) "A person who is a response action contractor with respect
    to any release or threatened release of a hazardous substance
    or pollutant or contaminant from a vessel or facility shall not
    be liable under this title or under any other Federal law to any
    person for injuries, costs, damages, expenses, or other liability
    (including but not limited to claims for indemnification or con-
    tribution and claims by third parties for death, personal injury,
    illness or loss of or damage to property or economic loss) which
    results from such release or threatened release."
(2) "Paragraph (1) shall not apply in the case of a release that is
    caused by conduct of the response action contractor which is
    negligent, grossly negligent, or which constitutes intentional
    misconduct."
   Based on a review of the legislative history of Section 1 19, the
Congressional rationale behind Section  119(a) can be interpreted
to be threefold.

• A  federal liability  standard of negligence  provides adequate
   performance incentives for  RACs working in the Superfund
   program
• A  federal negligence  standard holds RACs to the traditional
   "standard of care" that is applied  to  non-hazardous waste
   engineering, design and construction work
• A  negligence standard for RACs should provide prospective
   incentive  to the P&C insurance industry to develop liability
   insurance coverages for RACs active in the Superfund program

   During Superfund reauthorization, several  key Congressional
committees  reviewed the inclusion of a negligence standard for
RACs in various House and Senate Superfund reauthorization bills.
The following sections briefly review how each Congressional com-
mittee  viewed  the negligence standard for  RACs during the
reauthorization debate and how the final Section 119 negligence
standard language was adopted by Congress.

ENERGY AND COMMERCE COMMITTEE OF THE
HOUSE OF REPRESENTATIVES
   In  its report on RAc provisions (then designated Section 1 18)
of H.R. 2817, The Energy and Commerce Committee of the House
of Representatives noted the following regarding RAC liability
under CERCLA and the need for a preemptive negligence standard:
   "This section clarifies and differentiates the liability of response-
action contractors hired by EPA, other Federal agencies, the States
or persons undertaking cleanup of their own sites from the liability
of responsible parties for Superfund-related activities. Response-
action contractors currently face two distinct liabilities— liability
for future cleanup costs under the present Superfund law and third-
party liability under the various  State laws.
   Under the current Superfund law, response-action contractors,
despite  the exercise of due care and satisfactory performance of
contract specifications, could, by virtue of their involvement with
hazardous waste  sites, be included in the definition of owner,
generator or transporter and therefore could be subject to liability
under Superfund.
   Liability for third-party damages under State law adds additional
risks  which continue years after contract completion, potentially
exceeding both the contract amount and  the company's net worth.
Of particular concern is the potential application of strict or ab-
solute liability;-em;that is without regard to fault or willfulness— for
these damages.
   The section exempts response-action contractors from liability
under section 106 or 107 of CERCLA, or any other provision of
Federal or State law, for all costs, damages or any other claims
arising out of a release unless the contractor's negligent, reckless
or willful misconduct caused the release. The protections of this
section extend to any persons retained or hired by the response-
action contractor to perform services relating to the response action
under CERCLA.
  The section limits the coverage of this exemption from certain
liability provisions only to response-action contractors who would
not otherwise be liable under CERCLA if they had not entered
into a contract or agreement to perform a response action."*"
  The Energy and Commerce Committee supported the adoption
of a federally preemptive negligence standard and the provision
of discretionary  federal  indemnification  for pollution liability
arising from RAC negligence.

JUDICIARY COMMITTEE OF THE HOUSE
OF REPRESENTATIVES
   In its report on RAC provisions (designated Section 118), of H.R.
2817, the Judiciary Committee of the House of Representatives
noted the following regarding RAC liability under CERCLA and
the need for a preemptive negligence standard:
     "This provision eliminates strict, joint and several liability
   for response action contractors. Instead, these contractors will
   be liable only if their actions at a site are negligent or grossly
   negligent, or they constitute intentional misconduct. This  limi-
   tation on liability was included in order to ensure that companies
   will be willing to perform clean-ups in the future, and in order
   to give a reasonable incentive for insurers to provide prospec-
   tive insurance based on the traditional standard of performance
   that insurers cover (i.e. negligence).7"
The Judiciary Committee noted further that H.R. 2817 addressed
the two major problems  faced by  the current liability insurance
shortage.
     "First, it provides  a reasonable incentive for  insurers to
   provide prospective insurance to RACs. Second, it recognizes
   that regardless of the liability standard, it will be some  time
   before insurers can provide adequate insurance, and, therefore,
   it provides an interim form of protection to keep the Super-
   fund clean-up program functioning until insurers reenter the
   market.  . .  .It is important to understand that the principal
   risks  normally covered by liability insurance are to  defend
   against claims alleging negligence, and, if negligence is proven,
   to satisfy such claims. Put more simply, liability insurance covers
   negligence.8"
   The Judiciary Committee supported the adoption of a federally
preemptive RAC negligence standard as the first step towards
creating a statutory legal  climate for RACs in which insurers have
an opportunity to  operate under fundamental and traditional
insurance principles. In addition,  the Judiciary Committee sup-
ported the provision of discretionary federal indemnification for
pollution liability arising from RAC negligence.

PUBLIC WORKS AND TRANSPORTATION COMMITTEE OF
THE HOUSE OF REPRESENTATIVES
   In its report on RAC provisions (designated Section 124) of H.R.
2817, the Public Works and Transportation Committee of the
House  of Representatives noted the  following  regarding  RAC
liability under CERCLA and the need for a preemptive negligence
standard:

      "Under the concept of strict, joint and several liability which
    is applied to hazardous waste sites, any response action con-
    tractor working at that site is potentially liable for all removal
    and remedial costs associated with a release or threatened release
    of a hazardous substance from the site. This is so even though
    the contractor is following the requirements set forth in its  agree-
36    LIABILITY

-------
   ment with the Administrator, another Federal agency, a state
   or political subdivision or a potentially responsible party. The
   imposition of such liability is not appropriate in these cases,
   and for that reason, section 124 exempts response action con-
   tractors  from liability.9"
  The Public Works and Transportation Committee supported the
adoption of a federally non-preemptive RAC negligence standard
and the provision of discretionary federal indemnification for pol-
lution liability arising from RAC negligence.

SENATE BILL 51/REDESIGNATED H.R. 2005
  The Senate, in its version of the Superfund reauthorization bill
(H.R. 2005), supported a non-preemptive negligence standard for
RACs. The Senate amendment sought to amend the definition of
"owner or operator" contained in Section 101(20) of CERCLA.
The definition of "owner or operator" would have been modified
by the Senate to exclude RACs from liability under CERCLA
except to the extent that there is a release "primarily caused  by
the activities of such person." This modification in effect created
a federal negligence standard for RACs under CERCLA.
  The Senate supported the adoption of a federally non-preemptive
RAC negligence standard and the provision of discretionary federal
indemnification for pollution liability arising from RAC negligence.
CONFERENCE COMMITTEE REPORT
  In adopting final Section 119 language for SARA,  Congress
merged H.R. 2817 into Senate Bill H.R.  2005. For the purpose
of the RAC liability standard under SARA, Congress decided to
adopt the House Amendments with modifications. Congress, after
debating the options, was unwilling to impose a preemptive feder-
al negligence standard for RACs on the states or to provide federal
indemnification for RAC liability arising from a  strict liability
standard  at the state level.
  In its draft conference committee report on RAC liability and
indemnification provisions, Congress noted the following  regarding
RAC liability under CERCLA and the adoption of a non-
preemptive negligence standard:
     "Conference substitute—The substitute adopts the House
   amendment with modifications.
     The first modification to new subsection 119 provides that
   response action contractors shall not be liable except for their
   own negligence under CERCLA or any other Federal law for
   injuries, costs,  damages, expenses, or  other liability for any
   release or threatened release of a hazardous substance, pollu-
   tant or contaminant with respect to which it is a response action
   contractor. However, this section does not affect the liability
   of any person under any warranty under Federal,  State,  or
                                                           Table 1
                                                State RAC Liability Standards and
                                                   Other Relevant Information
Stale Statutory Strict
Liability for GTDs*
of Hazardous Waste
Alabama N
Alaska ฉ
Arizona N
Arkansas N
California CO
Colorado
Connecticut
Delaware
Florida
Georgia
Hawaii
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
Montana
Nebraska
Nevada
New Hampshire
New Jersey
New Mexico
New York
North Carolina
North Dakota
Ohio
Oklahoma
Oregon
Pennsylvania
Rhode Island
South Carolina
South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
Washington
West Virginia
Wisconsin
Wyoming
N
N
I
N
N
N
(Y)
N
N
N
N
N
N
N
N
N
Statutory Strict
Liability for
RACs
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
Statutory
Negligence
for RAC.
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
&
N
Y
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
Statutory
Limit on
Liability
N
N
N
N
N
Z Z Z Z Z
N
N
N
N
N
z z z z z
N
N
N
A
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
Common
Law Standard
for RACs
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
Z Z Z Z Z
z z z z z
Pending Applicable Ccmments/Nolel
Legislation Bill or on State Bill
Affecting RACl Statute f or Statute
N
N
N
N
N
N
N
N
Z Z Z Z Z
N
N
N
N
N
N
N
N
N
N
N
N
N
N
z z z z z
N
N
N
N
N
z z z z z
N
N
N
KB. 162 Negligence Standard for RACs


H.B-5071 RAC Indemnification by State




H.B.434 Negligence Standard, for RAO
Indemnification by Stale
                                                                                                         LIABILITY     37

-------
   common law.
     For purposes of Federal law, and  in  recognition of the
   inability of contractors to obtain insurance in the current market
   as well as their essential role in responding to releases caused
   by others, the conference substitute provides a standard of lia-
   bility based on negligence. Liability which might arise under
   non-Federal laws, however, is untouched by the conference sub-
   stitute. The existing standard of liability for responsible parties
   under CERCLA is maintained. The conference hopes that this
   amendment will  induce  States to deal with the  question of
   liability within their own borders. The conferees urge States to
   take note  of the Federal standards and review their own
   standards  of liability.10"
   While  Congress,  in enacting  Section  119, was  unwilling to
 preempt state statutes under which RACs potentially could be held
 to a standard of strict, joint and several liability, Congress was
 aware that not preempting such statutes placed the burden of the
 issue on  states Congress stated—
    "In passing a federal negligence standard  for RACs, the Con-
    ferees decided not to pre-empt  state law in this area, which may
   differ from the federal standard established by this Act. In so
   doing, the Conferees call special attention to the unique posi-
   tion of Response Action Contractors in the Superfund cleanup
   process. RACs are not the cause of the problem of hazardous
   waste sites, but their services are pivotal in carrying out cleanups.
   We accordingly urge states to take note  of this  new federal
   standard and review their standards of liability in this context
   so as to allow orderly progress in the cleanup of hazardous waste
   sites.""

 STATE LIABILITY STANDARDS FOR
 RESPONSE ACTION CONTRACTORS
   While the message of the Statement of Managers strongly urged
 states to enact negligence standards for RACs, only three states,
 New Jersey, Louisiana and New York, have enacted statutory negli-
 gence standards. While none of the remaining 47 states have enacted
 a statutory strict liability standard for RACs, the lack of a statu-
 torily defined negligence standard leaves RACs and commercial
 insurers in much the same state of uncertainty as before Section
 119 was passed. Under federal law, RACs are  held to a negligence
 standard for their response action activities; however, it is not clear
 whether they will be held to a strict liability standard or  a negli-
 gence standard under state law.
   An  increasing number of  states are  considering legislation
 affecting  RACs, including enacting a negligence  standard, pro-
 viding indemnification, or limiting the liability of RACs to a specific
 dollar amount  (Table I). However,  as indicated by Table 1,24 states
 possess statutes which are similar to CERCLA. These statutes hold
 hazardous waste generators, transporters and disposers to a strict,
 joint and several liability standard. Insurers  and RACs contend
 that  RACs potentially could be held strictly, jointly and severally
 liable under such state Superfund  statutes or under common law.
 Insurers and RACs fear a state court may potentially hold them
 strictly, jointly and severally liable in the 47 states without a specific
 statutory liability standard. To date, no state legal cases addressing
 the issue of the standard of liability for RACs have been published.

NEGLIGENCE STANDARDS FOR RACS  IN THE
STATES OF NEW JERSEY,  LOUISIANA AND NEW YORK
New Jersey
  Two states enacted negligence standards for RACs before the
passage of SARA, Section 119. New Jersey's negligence standard
(N.J. Rev. Stat. 58:10-23.1 Igl) was enacted in January,  1986 in
response to conditions similar to those that prompted the passage
of Section 119  of SARA. New Jersey contains  approximately 13"%
of the sites on the NPL. State Senator Paul Centillo, sponsor of
 the bill in the  New Jersey State Legislature, described the situa-
 tion  in New  Jersey  before  passage of  the bill. "Nothing was
                                                             happening in toxic waste cleanup, primarily because no one was
                                                             doing  the job. They could  not get insurance.  And without
                                                             insurance, they were betting their entire company every time they
                                                             went out  to do a job.12"
                                                               New Jersey's statute sets negligence as the standard of liability
                                                             for any persons performing hazardous  discharge mitigation or
                                                             cleanup services which results in injury to persons or property. In
                                                             addition, the statute creates a rebuttable presumption that the acts
                                                             or omissions were not negligent if the mitigation or cleanup services
                                                             adhered to generally accepted practices and state-of-the-art scien-
                                                             tific knowledge and utilized the best technology reasonably avail-
                                                             able at the time the services were performed.
                                                               The statement of  the New  Jersey State Senate Energy and
                                                             Environment Committee gave the reasoning behind the bill.
                                                                  "Senate Bill No. 3206 would limit the liability of persons en-
                                                                gaged  in hazardous waste site clean-up activities for injuries
                                                                related to these clean-up activities only to acts or omissions
                                                                which  can be shown to have  been  negligent. Under the provi-
                                                                sion of the New Jersey "Spill Compensation and Control Act,"
                                                                persons associated with a hazardous discharge,  including con-
                                                                tractors and engineers  hired to  clean-up  or mitigate the
                                                                discharge, are strictly, jointly and severally liable,  without regard
                                                                to  fault, for all damages resulting from the discharge. This
                                                                virtually unlimited liability has made  it almost  impossible for
                                                                hazardous  waste site clean-up contractors to obtain liability
                                                                insurance for their work, which in turn is hampering the state's
                                                                hazardous waste site clean-up program. By limiting the liability
                                                                of hazardous waste site clean-up contractors to their own negli-
                                                                gent acts or omissions, this bill would make it possible for con-
                                                                tractors to obtain liability  insurance  and  thus  continue their
                                                                clean-up activities. The committee amended Senate Bill No. 3206
                                                                to provide that  the provisions of this bill apply to contracts for
                                                                hazardous discharge mitigation or clean-up services entered into
                                                                before  this  bill and in  process  on the effective date of this
                                                                bill.13"

                                                             Louisiana
                                                               The Louisiana  negligence  standard  (La.  Rev.  Stat.  Ann.
                                                             9:2800.1) was enacted for  reasons similar to New Jersey's. The
                                                             Louisiana bill  creates a negligence standard  for  "architects or
                                                             engineers contracting to design or supervise hazardous waste miti-
                                                             gation, abatement or cleanup services. .  . ." Committee debates
                                                             in the  State Senate Judiciary Committee centered around the lack
                                                             of  professional liability  insurance  for architects  engaged in
                                                             hazardous waste mitigation work. State legislators stressed the fact
                                                             that the risk to public health was not created by those trying to
                                                             clean up the hazardous waste. Louisiana legislators  anticipated the
                                                             importance of state liability standards under SARA and wanted
                                                             to create a legal climate in the state where contractors and insurers
                                                             could  be certain of a standard of liability that they would be held
                                                             to.14

                                                             New York
                                                               New York is the only state that has enacted a statutory negli-
                                                             gence standard for RACs since the passage of Section  119. New
                                                             York's law (S.6448-A), enacted in August, 1987, limits the liabDity
                                                             of RACs working for the state to acts which are found to be negli-
                                                             gent, grossly negligent or involve willful and wanton misconduct.
                                                             The law also limits RAC liability to injuries or property damage
                                                             arising from acts or  omissions done in the course  of performing
                                                             their duties, not for the acts of others done before or after the
                                                             RAC's work. The new law  also eliminates joint and several lia-
                                                             bility for  RACs for property  damage and non-economic losses
                                                             relating to personal injuries if the RAC is  less than 50% liable and
                                                             did not act recklessly. The new law does  not apply to RACs who
                                                             are responsible for the spill, discharge  or inactive hazardous waste
                                                             site they are remedying.
                                                               New York's reasons for  enacting  this law  are  similar to the
                                                             reasons for enacting Section 119 of SARA. A memorandum in sup-
38
LIABILITY

-------
port of the law gives the New York Legislature's rationale for
enacting the law:
     "RACs in these areas provide a public service, and have
   recently found it difficult to obtain liability insurance at a
   reasonable premium or even at any price. The proposed legis-
   lation establishes several limitations on the liability of such con-
   tractors.  These  provisions should increase the number of
   contractors who work in the hazardous waste remediation field
   and should make such contractors more attractive insurance
   prospects. Under the bill, these contractors will be liable only
   for negligence, gross negligence, or willful or wanton miscon-
   duct, making the State's standard of liability for such contrac-
   tors like the standard of the Federal government established in
   the 1986 Federal Superfund Amendments.15"
  According to New York legislative officials, the driving force
behind the enactment of S.6448-A was the chronic problem of con-
tractors leaving the cleanup market because they could not get
environmental insurance.  In some cases, the state had to delay
hiring contractors in 1986 because of insurance problems, and offi-
cials feared the state eventually would become liable. New York
legislative officials also indicated that the lack of contractors is
the major reason why the State Department of Environmental Con-
servation is 2 years behind in its  cleanup schedule.16

CURRENT AVAILABILITY OF POLLUTION
LIABILITY INSURANCE FOR  RACS
  As of the writing of this article, commercial pollution liability
insurance still generally is not available for RACs  working on
Superfund sites in the 47 states with no defined liability standard
for RACs. At least one insurer, American International  Group
(AIG),  is offering pollution liability coverage to RACs working
on  Superfund sites in one of the states with a defined negligence
standard.
  AIG, a large commercial insurance company, has a program for
cleanup contractors providing third-party liability coverage for
bodily injury and property damage. The program covers contrac-
tual liability, limited completed operations and legal defense costs.
It is a claims-made policy with maximum limits of $1 million per
occurrence, $2  million annual aggregate, with a minimum  self-
insured retention of $50,000 and a minimum premium of $25,000.
  AIG  is willing to offer, on a site-specific basis,  coverage to
Superfund RACs in states with a statutory negligence standard.
However, to date, no Superfund contractor has purchased this
coverage. Insurance brokers for AIG have not seen a great  deal
of enthusiasm for this coverage on the part of RACs. Since the
majority of large Superfund contracts are awarded on a zone basis,
contractors do not see it as worthwhile to purchase insurance for
work in only one of the many states in which they perform response
actions.
  In response to the lack of pollution liability insurance for RACs,
a captive insurance company, DEMETER, has been formed by
the Hazardous Waste Action Coalition (H WAC) of the American
Consulting Engineers Council (ACEC). DEMETER provides
professional and general liability coverage on a claims-made basis
to  qualified  members  with  a  hazardous  waste  practice.
DEMETER's coverage will include coverage for pollution claims
otherwise excluded from a member's professional liability or com-
prehensive general liability policy. DEMETER's  management
stresses the fact  that a captive  insurance company, such as
DEMETER, should be viewed as temporary relief and a  substi-
tute mechanism formed to ease the shortage of commercial pollu-
tion liability insurance.17
  Section 119(c)(4) of CERCLA requires RACs to obtain com-
mercial insurance if it is available at a "fair and reasonable price"
before the U.S. EPA will indemnify the RAC. The statute  also
requires, in the case of a contract covering more than one facility,
that the RAC agree to continue to make diligent efforts to obtain
commercial pollution liability insurance each time the RAC begins
work under a multi-site contract at a new facility.
  At the writing of this article, the U.S. EPA has not determined
whether this AIG policy is acceptable for RAC coverage. However,
if the U.S. EPA does determine that this policy or others similar
to it are commercially available to RACs at a fair and reasonable
price, RACs will be required to purchase this insurance and U.S.
EPA indemnification will be reduced by the amount of the insur-
ance coverage.
CONCLUSION
  The enactment of a federal negligence standard for RACs by
Congress in Section 119 of SARA is the first major step toward
providing the P&C insurance industry with incentive to create a
viable and stable commercial insurance market for RACs partici-
pating in the Superfund cleanup program. A federal negligence
standard for RACs helps to create a statutory legal climate for
RACs in which insurers have an opportunity to operate under fun-
damental and traditional insurance principles. However, because
the federal negligence standard for RACs in Section 119 of SARA
does not preempt state law, insurers still may be reluctant to offer
insurance coverage to RACs participating in the Superfund clean-
up program because of the potential that RACs may be held to
a strict, joint and several liability standard under state laws similar
to CERCLA.
  The enactment of a negligence standard for RACs by the states
of New  Jersey, Louisiana and New York  is one factor that may
lead to the availability of pollution liability insurance for RACs
in those states because it further clarifies the legal standard that
RACs will be subject to under state law. Once it is clear that RACs
and their insurers will not be held to a standard of strict, joint and
several liability under state law for response activities, insurers seem
more willing to insure RACs working in Superfund program. As
the state legal climate for RACs becomes more certain and as the
P&C insurance  industry restores financial capacity and more
thoroughly understands the RAC liability  exposure  related to
Superfund program, the U.S. EPA fully anticipates that its role
in providing Section 119 indemnification will diminish and will be
replaced by the private sector.
REFERENCES
 1 . Superfund Amendments and Reauthorization Act of 1986, Section
 2. Black's Law Dictionary 930 (5th ed. 1979).
 3. Black's Law Dictionary 1275 (5th ed. 1979).
 4. Black's Law Dictionary 951 (5th ed. 1979).
 5. Report of Public Works and Transportation Committee of the House
   of Representatives Regarding H.R. 2817, 1986, 92.
 6. Report of the Energy & Commerce Committee  of the House of
   Representatives Regarding the Response Action Contractor Provision
   of H.R. 2817, 1986, 92.
 7. Report of the Judiciary Committee of the House of Representatives
   Regarding the Response Action Contractor Provision of H.R. 2817,
   1986, 26-28.
 8. Report of the Judiciary Committee of the House of Representatives
   Regarding the Response Action Contractor Provision of H.R 2817
   1986, 29.
 9. Report of the Public Works and Transportation Committee of the
   House of Regarding the Response Action Contractor Provision of H R
   2817, 1986.

10. Draft Conference Committee Report on the Superfund Amendments
   and Reauthorization Act, 1986, 53.
1 1 . Statement of Managers Accompanying SARA Conference Bill, 1986.
                                                                                                            LIABILITY    39

-------
  12. Newark Star Ledger; Jan. 17, 1986.                                    15.  New York Governor's Program Bill Memorandum on S.6448-A, 1987,
  13. New Jersey State Senate, Energy and Environment Committee, State-             2-
      ment to Senate. No. 3206, Dec. 5,  1985.                                ,6  Hazardous Matcria, Con(ro| Rescarch ,nstilute, FocttSi Aug. 1987i 2
  14. Legislative History Tapes Concerning RAC Negligence Standard for
      the State of Louisiana, 1987.                                          17.  Business Insurance, July 6, 1987.
40     LIABILITY

-------
                 Calculating a  Risk  Premium for a  CERCLA Site

                                                  Yardena Mansoor
                                          Planning Research  Corporation
                                                  McLean, Virginia
                                                    Thomas  Gillis
                                     U.S. Environmental Protection Agency
                                     Office of Waste Programs Enforcement
                                                  Washington, D.C.
ABSTRACT
  Interim CERCLA Settlement Policy allows the U.S. EPA to
offer a PRP limited release from liability for cleanup costs in
exchange for a present sum certain that includes a risk premium.
"In limited circumstances, the release may become effective upon
payment for Federal Government cleanup, if the payment includes
a carefully calculated premium or other financial instrument that
adequately insures the Federal Government against [the uncertainty
that  the remedial response will cost more than was estimated at
the time of settlement]."
  This paper discusses the principles applicable to setting an
appropriate level for the risk premium. The risk premium calcula-
tions are  based on  the cleanup cost estimates in a site's FS and
ROD, and on the U.S. EPA negotiator's estimates of the uncer-
tainties that the PRPs would transfer to the U.S. EPA as a result
of the cash out agreement.  The risk premium therefore depends
on the negotiator's estimation of the precision of the ROD cost
estimates and on the negotiator's assumption (in the absence of
complete  information) about the probability  distribution  for
cleanup cost.

DISCLAIMER
  The views expressed in this paper are those of the authors alone
and  do not necessarily reflect those of the U.S. EPA.

INTRODUCTION
  Under CERCLA, as amended by SARA, responsible parties are
liable for  the costs of cleanup of a hazardous waste site.  Their lia-
bility is based on the existence of relationships linking them to the
contaminated site or to the hazardous  substances at the site. Under
CERCLA, current owners of a facility, owners of the facility at
the time of disposal and all operators,  transporters, generators and
persons who "otherwise arranged for disposal or treatment" are
liable for  the costs of cleanup. CERCLA does not mandate a spe-
cific  standard of liability; the courts have concluded that strict
liability and joint and several liability are the applicable standards.
This  implies that the government may recover the entire cost of
cleanup from any liable party without needing to identify all respon-
sible  parties and apportion costs to the parties on the basis of their
contribution to the harm1.
  In  practice, equity and efficiency require the allocation of cleanup
costs among the PRPs, which number in the hundreds  for some
sites. Under Section 122(e)(3)(A) of SARA, the U.S. EPA may
at its discretion develop a Non-Binding Allocation of Responsi-
bility (NEAR) which allocates percentages of total response costs
at a  facility to PRPs. The purpose of the NEAR is to expedite
settlement of cost sharing for the cleanup,  but it is not binding
on the PRPs or on the government.  The NEAR allocates 100%
of the response costs among the PRPs.
  In preparing its allocation of responsibility for cleanup costs,
the U.S. EPA  may consider many factors:
• The nature of the wastes with which each PRP was involved,
  including
  - Volume
  - Toxicity
    Mobility
• Settlement policy criteria, including
  - The strength of the evidence tracing the wastes at a site to each
    PRP
  - The ability  of each PRP to pay
  - Litigative risks in proceeding to trial
  - Public interest considerations
  - Precedential value
  - The value of obtaining a present sum certain
  - Inequities and aggravating factors
  - Nature of the case that remains after settlement

  The U.S. EPA feels that NBARs are particularly appropriate
where there are a sizable number of de minimis contributors6. In
negotiating with PRPs for a site cleanup, the U.S. EPA may offer
a cash out settlement to certain PRPs responsible for low volumes
of low toxicity waste. This conserves administrative resources for
negotiating with PRPs who will make substantial contributions to
the cleanup costs.
  The U.S. EPA may also consider a cash out settlement with a
PRP that is not a de minimis contributor. This PRP would be con-
sidered for limited release from liability in exchange for a present
sum certain that includes a risk premium. The U.S. EPA Interim
CERCLA Settlement Policy states, "In limited  circumstances, the
release may become effective upon payment for Federal Govern-
ment cleanup, if the payment includes a carefully calculated premi-
um or other financial instrument that  adequately insures the Federal
Government against these uncertainties [of the  actual costs of
cleanup]."5 That is, a cash out arrangement is intended to  pro-
vide the U.S. EPA with a payment that adequately protects against
the risk that the remedial response will cost more than was  esti-
mated at the time of settlement.
   A separate type of risk is not addressed by  the risk premium:
the risk that at  a future date the remedy will  prove inadequate.
This risk is covered in the settlement agreement by "reopeners"
in the settlement agreement that allow the Government to: (1)
modify terms and conditions of the release from liability and (2)
recover  additional costs. At a minimum,  these circumstances
include imminent and substantial endangerment to public health,
welfare or the environment posed by:

•  Previously unknown or undetected conditions arising or dis-
   covered at the site after the time of the agreement
•  Additional information not available at the time of agreement
   concerning scientific determinations on which the settlement was
   premised (for example, health effects associated with levels of
   exposure, toxicity of hazardous substances and the appropriate-
                                                                                                       LIABILITY    41

-------
  ness of the remedial technologies for conditions at the site)

  A risk premium, in this analysis, is defined to be the compensa-
tion to the U.S. EPA for taking the risk of the selected remedy
cost being higher than the cost expected or estimated at the time
of settlement with the PRP= The total premium paid by the PRP
would consist of the allocated share of the previous government
expenditure at the site, the allocated share of the response cost as
estimated in the ROD and a risk premium. (To facilitate the dis-
cussion that follows, we do not repeat "allocated share of" in every
reference to the PRP's payment of estimated cost. It therefore
appears that only one PRP is  paying the total estimated cost and
a risk premium based on that cost. A workable approach is to calcu-
late a total risk premium and allocate it among PRPs using the
same allocation factors as the NBARs.)
   The concept of risk premium as compensation for accepting a
risk is well-established in the fields of finance, economics and
insurance when a person or entity liable for a risk (a PRP) transfers
the risk to another  entity (U.S. EPA). Although the concept of
a risk premium has been included in several past U.S. EPA Super-
fund negotiations, there currently are no guidelines on how to set
a  premium level. A risk premium can be  calculated by various
methods. This paper explores one approach to setting an adequate
"target" risk premium for any chosen level of financial risk. Of
course, the actual  risk premium will depend on negotiations
between the U.S. EPA and the PRP.

THE CASH OUT AGREEMENT
   In  accepting a risk premium for limited release from liability,
the U.S. EPA assumes the risk that if the actual costs of the
remedial response are higher than expected, the total pool of
recovered costs would be less  than in the absence of the cash out
agreement. When all the PRPs at a site are released, the U.S. EPA
assumes the risk that the total remedial response cost may exceed
the total premium collected.
   The use of cash out payments will affect the transactions shown
in Table  1, which  shows the cost  and  financial  risk transfers
resulting from the U.S. EPA acceptance of a PRP cash out.
                           Table 1
         Cost Risk Transfers Between PRP and U.S. EPA
                  Resulting From a Cash Out
  ESTIMATED
  COST OF
  REMEDY
  RISK OF
  COST
  OVERRUNS
  TRANSACTION
  COSTS
                    Without Cnh Oul
PRP pays for cleanup cost
incurred tl the site
(past ceruin and future
enjoined)
PRP bun coil overrun
from estimated clean up
cost (unknown)
Trinuction costi
for PRP tnd EPA
    With Cnh Out


PRP payment to EPA for
eipected cleanup cost
(certain)
PRP payl risk premium to
EPA for potential overruns
(certain)
Lower trtniaction coiti
for PRP ind EPA
  The PRP benefits from a cash out arrangement in several ways.
By obtaining a limited release, the PRP reduces its uncertainty of
increased future costs of the selected remedial response. The other
principal benefit is the savings in transaction cost, including legal
fees, internal staff resources and other expenses related to a con-
tinuing legal action.
  Like the PRP, the U.S. EPA can benefit from the cash out ar-
rangement by increasing the likelihood of collecting the total premi-
um payment and by lowering the transaction costs. The sum certain
payment reduces the risk  that the U.S. EPA may not collect the
expected cleanup cost through legal action, due to either the ina-
bility of the PRPs to pay or weakness of the case. The reductions
in transaction costs result from: (1) savings the resources that would
be spent on extended negotiation and litigation, and (2) reduction
in the number of PRPs remaining, the U.S. EPA enhances the
chance of negotiating a settlement with them.
  Therefore, the benefits from a cash out arrangement may be sub-
stantial, particularly at sites with a large number of PRPs. The
arrangement may result in significant savings, particularly of trans-
action costs,  to all parties.

APPROACHES TO SETTING A RISK PREMIUM
  Setting a risk premium involves two elements: (1) estimating the
magnitude of the risk (the probability distribution of actual costs
around an estimated cost) and (2) determining the relative willing-
ness of the risk seller (the PRP) and the risk buyer (U.S. EPA)
to accept the risk of the actual remedy cost exceeding the estimated
cost. Because of differences in their estimates of the variance of
future costs or in their willingness to accept risk, the parties to the
risk transfer transaction may have different estimates of the risk
premium appropriate to compensate for the transfer of risk.
Expected Cost and Its Variance
  Information needed to set a risk premium comes primarily from
the FS. The  FS provides the basic information regarding  the
remedies available for the site, their expected effectiveness and their
estimated  costs. This information also is contained in the ROD
for the selection of remedy. The information  provided in these
sources, however, is  typically a point  estimate for cleanup cost.
But before the remedy is implemented, cost is a random variable
with a mean value of the estimated cost. Since the risk premium
addresses  the chance that the actual cleanup cost will exceed the
expected cost, additional information is needed about the proba-
ble distribution of costs around the estimated cost.
  Some FS or ROD cost analyses present remedial cost estimates
as a range; that is, the remedy is estimated to cost X ฑ Y%. This
statement may be interpreted as a confidence interval; for example,
there is a 95% likelihood that the cost will fall within Y% of X.
But this last statement is an interpretation, not a statistical deduc-
tion. It expresses the  U.S. EPA's (or the PRP's) confidence in the
accuracy of the FS or ROD cost estimates as developed by the
engineering contractors.

Expected Cost Distribution Approach
  One analytical approach uses  standard  probability  density
functions to represent hypothetical distributions of cleanup costs
around the mean of an estimated cleanup cost. Hypothetical dis-
tributions often are used by the insurance industry to assist in quan-
tifying risk when  little actuarial data on loss are available. This
method also is useful for quantifying a risk premium, since data
on the cost of cleaning up uncontrolled hazardous waste sites are
scarce.
  Assuming a distribution of costs allows the U.S. EPA to estimate
an adequate risk premium to cover any given probability that actual
cleanup costs will be greater than available financial resources. That
is, it answers the question, "How large must a risk premium be
to have only a Z% chance of the actual cleanup cost exceeding
the premium?" This methodology does not produce a unique risk
premium, but calculates a number of premiums as a function of
alternative levels of  acceptable risk.
   Given a distribution function for the cleanup cost, a total pre-
mium (expected cost plus  risk premium) can be calculated based
on one of several underlying principles2:

•  Net premium principle: Setting a total premium payment that
   equals the  expected mean of the cost distribution:
     P =  E(C)                                           (1)
where P is the total premium and E(C) the expected value of C,
the cleanup cost size. The risk premium, in effect, is zero.

Expected value principle: Setting a total premium proportional to
the expected  value of the cleanup cost:
     P =  (1  + a) *  E(C)                                 (2)
42    LIABILITY

-------
where a is greater than 0. The risk premium equals [a * E(C)]

Variance principle: Setting a total premium equal to expected cost
plus a risk premium proportional to the variance of C
  P = E(C) +  b * Var(C)                                (3)
where b is greater than 0.

Maximum loss principle: Setting the total premium to the maxi-
mum possible cost:
  P = max (C)                                           (4)

  The first and the last principles for calculating the premium
represent, respectively, the extreme cases of receiving no compen-
sation for taking the risk and receiving so much compensation that
any risk taking is avoided. It would be illogical for the risk buyer
(the U.S. EPA) to accept the first arrangement or for the risk seller
(the PRP) to accept the latter. Intermediate approaches such as
the expected value principle and variance principle are more equita-
ble and more typical of actual applications. As will be shown below,
the expected value principle is associated with assuming an exponen-
tial distribution for cleanup costs and the variance principle is
associated with assuming a normal distribution for cleanup costs.
The alternative  assumed  distribution  functions can produce
significantly differing risk premiums for  any given  risk of cost
overrun.

CALCULATIONS BASED ON HYPOTHETICAL
DISTRIBUTION FUNCTIONS
  In response to limited information on the distribution of cleanup
costs, this analysis examines two alternative distributions of a cost
function. These alternatives are illustrated in Figure 1. One of these,
the normal distribution, assumes that cleanup costs are symmetri-
cally distributed above  and below an estimated cost. The other,
an exponential distribution with the same first or central moment
as the normal distribution, assumes a higher probability (relative
to the normal distribution) for extremely low cost (region A) and
for extremely high cost (region  B). These distribution functions
are used to address the question  "What is the probability that the
cleanup cost will exceed a given value?"
                             E(C)
                     Expected Cost - Mean of Distribution

                           Figure 1
             Two Standard Probability Distributions


  Each of two distribution functions determines a risk premium
dependent on risk acceptance level. The expected cost distribution
approach sets a risk premium appropriate for a given probability:

  Pr[C>P]                                               (5)

or the probability that the actual cleanup cost (C) will be greater
than the total premium (P), which in turn is defined as expected
cost plus a risk premium.  This analysis addresses  the question
"What should the total premium be for a given probability of final
cleanup cost  exceeding the total premium?"
  Using an exponential distribution for cleanup costs:

  Pr[C>P] =  1 - Pr[CP] =  e-KP                                      (9)

For the exponential distribution k = l/E(c), where E(c) is the ex-
pected value of c — that is, the remedial cost estimated in the FS
or ROD, one then obtains:
  Pr[C>P] =

Solving for P:
  P  = ln(Pr[C>P]) * E(c)
                                                        (10)
                                                        (11)
  For a given probability of the final cleanup cost exceeding the
total premium, the total premium multiplier is -ln(Pr[C > — ] and
the risk premium multiplier is the total premium multiplier minus
one. Table 2 column (A) shows the risk premium as a multiple of
estimated cost for several probabilities that total cleanup cost will
exceed the premium.
  A similar analysis can be performed assuming a normal distri-
bution for cleanup costs. In this case:

  Pr[C>P]  =  0.5 -the area under the standard normal  curve
                   between 0 and Z(P)                  (12)

where Z(P), the standardized variable, is equal to:

  ^^                                             03)
and a is the standard deviation of the cleanup cost. The term Z(P)
is the number of standard deviations from the mean that would
yield the desired probability that cleanup cost  would exceed the
sum of expected cost plus risk premium. Values  of Z can be found
in standard statistical tables for the normal density function. But
without data to empirically estimate the standard deviation, some
assumption is required here. For example, if the FS reports that
remedial costs may be as much as 30% higher than the  point esti-
mate and if one assumes that this represents a 95% confidence
interval or two standard deviations, then:
                                                                    (1  + 0.3) * E(c) = E(c) + 2 * a

                                                                    Solving for O

                                                                    a= 0.15  * E(c)
                                                        (14)
                                                        (15)
                                                                    Substituting in Equation (13) and evaluating for chosen levels of
                                                                    Pr[C > P] yields the risk premium multipliers shown in column (B)
                                                                    of:
                           Table 2
      Risk Premium as a Multiplier of Estimated Cleanup Cost
 Probability that Cleanup Cost
  Will Exceed Total Premium
           1.0
           2.5
           5.0
           10.0
           15.0
           20.0
                               (A)
                           Cleanup Costs
                           Exponentially
                            Distributed
                           (risk premium)

                              3.61
                              2.69
                              2.00
                              1.30
                              0.90
                              0.61
       (B)
   Cleanup Costs
Normally Distributed
95% confidence: 4.tyn,
 (Z) (risk premium)
 2.33
 1.96
 1.64
 1.28
 1.04
 0.84
0.35
0.29
0.25
0.19
0.16
0.13
                                                                                                               LIABILITY     43

-------
Choice of Distribution Function for
Setting a Risk Premium
  In the absence of empirical data on cleanup costs, a virtue of
this approach to setting a risk premium is that it reduces the
arbitrary quality of the decision by making the underlying assump-
tions more explicit. The choice of the assumed distribution function
for costs is important. This analysis has selected the normal and
exponential functions to calculate a relationship between the risk
premium and the probability of final cleanup cost exceeding the
total  premium.  A normally distributed cost curve  implies that
average or mean cost is also the one with the highest probability.
An exponentially distributed  cost curve implies that costs lower
than the mean have a higher probability than the mean cost, but
that costs significantly higher have a higher probability than under
a normal distribution. (Refer again to Figure 1). insurance com-
panies  often use  such  "long  tail"  distributions   for  low-
probability/high-consequence liabilities.
  The choice of distribution function has a dramatic effect on the
level  of risk premium.  Table 2 shows risk premiums for each
assumed distribution of cost function for selected probabilities that
the actual final cleanup costs will exceed the sum of allocated esti-
mated cost plus risk premium. For example, given an exponential
distribution, a risk premium of \30% of estimated cost will result
in a 10% chance that recovered costs will fall short of cleanup cost.
If costs are assumed to be normally distributed, however, a risk
premium of less than 20% of estimated cleanup costs will provide
the same level of risk compensation to the U.S. EPA  (given its
95 percent confidence level in the accuracy of the estimated cost).
  The results of this analysis  can be used in conjunction with the
U.S. EPA's expected long-term cleanup scenarios and its percep-
tion of the distribution of the cleanup costs. Results under the
normal distribution may be considered low for arriving at a risk
premium because cleanup costs at hazardous waste sites have been
historically underestimated. This distribution results from uncer-
tainties surrounding the remedial technologies and site characteris-
tics. A more  conservative approach  (from  the  U.S.  EPA's
perspective) would be to use the results of the exponential distri-
bution. However, the resulting large premium multiplier may dis-
courage the PRPs from a cash out arrangement; attaining a low
level of cost overrun risk would require the PRP to pay a risk
premium several times larger than the expected cost.
  Ultimately, the validity of the risk premium calculation depends
on defensible assumptions regarding the appropriateness of the
cleanup scenarios, the accuracy of the cost estimates and the distri-
butional form of the cleanup costs. As more experience is gained
in estimating cleanup costs and implementing  the remedy, the
assumptions used to quantify the risk premium should be refined
accordingly. Better estimates of expected remedial costs  should
result in lower risk premiums for  any given level of acceptable risk.
In the meantime, this framework can accommodate new informa-
tion as it is generated.

REFERENCES
I.  Garber, E. J., "Common Law of Contribution Under the 1986 CERCLA
   Amendments," Ecology Law Quarterly.  14, 2, 1987, 365-388.
2.  Gerber, H.U., "An Introduction  to Mathematical Risk Theory". S.S.
   Huebner Foundation for Insurance Education, University of Pennsyl-
   vania, Philadelphia, PA,  1979.
3.  Raiffa, H., Decision  Analysis.  Introductory Lectures on Choices
   Under Uncertainty. Addison-Wesley Publishing  Company, Reading,
   MA, Second Edition, 1970.
4.  U.S. EPA, Office of Emergency and  Remedial Response and Office
   of Waste  Programs Enforcement, Guidance on Feasibility  Studies
   under CERCLA. U.S. EPA/540/0-85/003. June 1987. Chapter 7.
5. U.S. EPA, "Hazardous Waste Enforcement Policy," Request for Public
   Comment. Federal Register. 50,  24. 1985. 5034-5044.
6.  U.S. EPA, 1987, Superfund Program; Non-Binding Preliminary Allo-
   cations of Responsibility (NBAR), (Request  for  Public Comment)
  Federal Register, 52. 1987, 19919-19921.
7.  Weisberg, H.I. and Tomberlin T.J.. "A Statistical Perspective on
   Actuarial Methods for Estimating Pure Premium from Cross-Classified
   Data," J. of Risk and Insurance. 1982. 539-563.
44    LIABILITY

-------
           Historical  Risk  Assessment of  Environmental  Liabilities
                                 At  Former  Industrial Properties
                                                     Sandy Peterson
                                              Risk  Science International
                                                   Washington,  DC
ABSTRACT
  This paper  discusses  the  Risk Science International  (RSI)
approach to evaluating  liability risks associated with former
industrial properties. Former industrial sites present risks  of
environmental impairment that can result in a wide range of serious
liabilities for subsequent property owners and investors. The poten-
tial liabilities can range from nuisance claims and legal violations
to multi-million dollar lawsuits and Superfund involvement.
  The contaminants that present ongoing concerns of environ-
mental impairment vary according to the nature and scale of activity
of the former industrial operations. Former industries often created
by, products that can pose ongoing environmental risks due equally
to past waste disposal practices and to the actual hazardous charac-
teristics of the wastes themselves. The history  of site use and the
site-specific characteristics  of the environmental setting are the
primary variable factors influencing the potential for environmental
impairment.
  RSI has developed an historical risk assessment methodology
to acquire  site-specific knowledge  of  potential environmental
impairment from former industrial sites and to identify the extent
of possible liabilities. The  historical  risk assessment serves as a
comprehensive risk management tool enabling informed decisions
to be made concerning necessary remedial investigations or actions.
It is an essential element of pro-active risk management strategies
that provide a company with the flexibility to address environ-
mental liabilities in the most efficient and cost-effective fashion.
Historical risk assessments should be  considered for any sites
suspected of being sufficiently contaminated to warrant investi-
gation or remediation under CERCLA or equivalent state rules.

INTRODUCTION
  This paper discusses Risk Science International's approach to
evaluating the potential risk of environmental liabilities associated
with former industrial properties. A growing  recognition of the
potential risks of environmental impairment created by past in-
dustrial activities has emerged over the past decade. A number of
these former operations created by-products that can pose ongoing
risks  of environmental impairment  due equally to past  waste
disposal practices and to the hazardous characteristics of the wastes
themselves. Each situation creates potential liabilities that can best
be faced through an effective risk management strategy based on
the knowledge afforded  by risk assessment.

ENVIRONMENTAL LIABILITIES
  Potential environmental liabilities related to former industrial
sites can range from chronic nuisance claims and minor legal viola-
tions to multi-million dollar lawsuits and Superfund involvement.
These liabilities can arise as the result of routine industrial activities
as well as the on-site and off-site disposal of wastes.
  Some past industrial sites and technologies that present a serious
potential for environmental impairment include town gas manufac-
turing plants, munitions plants and hat factories that used mercury.
Agricultural uses can result in impairment from pesticide and fuel
residues. Past activities as obscure as a roadside diner and garage,
or a popular local dump site, can pose impairment risks from aban-
doned storage tanks, buried wastes and other hidden exposures.
  The greatest potential liability that may be associated with former
industrial properties is  environmental impairment of sufficient
magnitude to present an imminent hazard to human health and
the environment. Depending on the extent of the contamination
and the location of the site, listing on the U.S. EPA's NPL under
Superfund could occur. In other instances, states or local authorities
could enforce remedial actions using Consent Orders based on a
variety of environmental statutes.  Cleanup levels ranging from
several ppm to negligible amounts in the ppb range may be enforced
for different contaminants by federal or state agencies. Certain
elaborate sampling and analytical protocols may be required. These
terms alone can predetermine a high magnitude of final costs for
a remedial investigation and action.
  Potential  groundwater contamination presents perhaps  the
greatest liability because of the extreme costs of effective ground-
water remediation. Soil, sediment and surface water cleanups can
be very widespread and costly. Generally, the costs for site remedial
investigation and cleanup can be greatly reduced if the problem
is addressed before state or federal regulatory agencies are forced
to take the lead. More favorable safe cleanup levels may in many
instances be negotiated when a company has the initiative to under-
take remedial actions, thereby freeing state resources for higher
priorities. Nevertheless, evolving legislation as  typified by  the
ECRA property transfer requirements now in effect in New Jersey
may drastically reduce  the flexibility  in  remedial approaches
presently allowed  under current  law.
  Many potential liabilities arise from the possibility of toxic torts
and other citizen actions. Nuisance and damage suits can be filed
on the basis of contaminated water supplies, loss of property use,
unhealthy air quality conditions and other concerns.  Once a
problem has fully  entered the public domain and become fueled
by publicity, liabilities can be greatly compounded in unforeseen
ways.
  Potential liabilities take on special importance during property
transactions.  In today's political and economic climate, good
management practices demand that both the purchaser and seller
                                                                                                         LIABILITY    45

-------
understand the current and potential future risks associated with
former industrial sites. A risk assessment can be used in many cases
to influence the terms of the transaction. For example, the pur-
chaser might reduce the  sale price, require corrective action or
receive indemnification. A seller can use the assessment to  docu-
ment the condition of the property at the time of sale to protect
against future claims alleging environmental impairment from past
site activities.

RESEARCH METHODOLOGY
  An historical risk assessment method is used to identify existing
and potential liabilities as well as areas of uncertainty that  could
suggest possible liabilities. The risk assessment process examines
the past industrial operations at the site, determines the potential
for environmental impairment associated with past activities and
evaluates possible liabilities associated with past and current site
use and conditions. The risk assessment also defines any areas of
uncertainty that could be clarified through additional research or
field investigations. Specific recommendations are provided to con-
duct focused remedial investigations that may be necessary to  quan-
tify the magnitude of the risks. An assessment goes beyond simple
risk identification to evaluate the potential consequences associated
with the risks.  The historical risk assessment thus serves as a com-
prehensive risk management tool on which to base decisions con-
cerning the need for and scope of any remedial efforts.
  Historical risk assessments are based on information obtained
from public sources and  through interviews with knowledgeable
past  and present employees  and local residents. The risk assess-
ment process is initiated by meeting with the company to define
the limits of the historical investigation. Permission is requested
to use various sources of information including libraries, historical
societies  and  interviews  with knowledgeable  people.  In  most
instances, investigations within the community will be conducted
under the guise of a general historical study. RSI has found this
approach to be very productive in gathering a wealth of informa-
tion  from respondents while protecting the company's needs for
confidentiality. In addition, the unlimited scope of the interviews
generally yields seemingly inconsequential information that may
prove valuable when pieced with other data.
  In RSI's experience, the historical risk assessment process can
be rendered more thorough and cost-effective by assessing two or
more sites at a  time that are grouped within a geographical region.
Research time is minimized by simultaneously obtaining informa-
tion on all sites from the various identified sources. This approach
generally helps to maintain client confidentiality by eliminating the
focus on a particular location  and substantiating the  premise of
a historical study. A detailed research phase is then initiated that
focuses on the  two most crucial aspects of the site assessment: the
environmental setting and the history of the property.

Environmental Setting
  A  site inspection is conducted to gather information on current
operations and all active or abandoned structures. The entire site
is examine for evidence of past or present contamination as well
as for any physical remains of the former industry. Information
is obtained on the environmental setting, including soils, drainage
patterns and off-site environmental characteristics. Land use
patterns of adjacent areas  also are determined during the site
inspection. Any observable environmental impairment is noted in
detail for subsequent discussion in the section of the  assessment
dealing  with  possible liabilities  associated with current site
conditions.
  The principal sources of environmental data employed are the
USGS state soil conservation and geological agencies and state en-
vironmental protection agencies. If necessary, well drillers and other
local sources may be contacted. Publicly available information is
obtained on characteristics of surface water, soils and groundwater
in the vicinity of the plant site. The subsurface investigation focuses
on soil  types,  surficial geology, local  hydrogeology and  local
groundwater use. The surface  investigation determines water
sources,  on-site water use and disposal, site drainage patterns,
receiving streams and local land use.
  Documentation of existing groundwater  and surface water
quality is also sought. All findings are compiled and analyzed to
determine the environmental sensitivity of the site with respect to
potential impairment from former operations. RSI integrates thiป
information with knowledge of the various materials handled or
disposed of on-site and the characteristic environmental fate of
these materials to determine the potential for migration through
different environmental routes. A similar exercise is performed in
the event that off-site waste disposal has occurred in a known or
suspected location.
Historical Research
  The principal  sources  of site-specific historical  information
include town and county libraries, historical societies, local public
works departments, governmental archives and knowledgeable in-
dividuals. With the permission of  the  company,  all of these
possibilities are exhausted in the search for historical data. Of par-
ticular interest are historic photographs, aerial photos, town plans
and the Sanborne Insurance Atlas. These can provide information
on site layout and possible operating changes over time. These data
are preliminarily analyzed to determine areas of relative certainty
and uncertainty concerning past  operations  and practices. The
analysis  forms a  basis for subsequent  in-depth investigations
through interviews.

Interviews
  The identification and interviewing of knowledgeable individuals
is an especially challenging task. In some instances, the company
may know of past or present employees who worked at the plant,
or relatives of such people. Elderly neighbors of the existing site
may have direct knowledge of the former operation or they may
know of individuals who  do. The personnel of the  local library
or historical society also may know of knowledgeable individuals
in the community. RSI has found from experience that, while these
individuals are  few, locating one knowledgeable person generally
leads to the identification of more such individuals or their direct
descendants.
  As discussed  earlier, the interviews are represented as being part
of an historical  study of former industrial operations in the region.
While the respondents may tend to suffer lapses in their recollec-
tion, they typically hold a strong interest in local history and an
eagerness to share their knowledge. Information sought in the
interview typically focuses on the history of ownership; site layout;
technologies  employed;  raw  materials, products and wastes;
shipping, storage and disposal practices; fires or other  unusual
occurrences;  chronic effluents or nuisance conditions;  and site
demolition.  Identified areas of uncertainty are carefully probed,
while established information is verified.
  As the interviews progress and the information coalesces, the
respondents may be recontacted to expand upon or verify certain
details.  It has been our experience  that the interviews generally
provide the greatest wealth of historical information, although
problems do arise in resolving conflicting accounts. Unfortunately,
we  have reached a point in time when the number of potential
respondents is becoming increasingly limited, and for certain sites
it may prove impossible to find them. It is RSI's contention that,
at this late date, time is of the essence in availing ourselves of these
diminishing sources of first-hand information. The historical record
is becoming more incomplete as time progresses, underscoring the
urgency of initiating any needed historical risk assessment in the
immediate future.

RISK EVALUATION
  Once all of the relevant  sources of information have  been ex-
hausted, the data are carefully compiled, analyzed and presented
in various summary forms. For example, site layouts are drawn
up if no accurate maps are available. Waste generation rates and
46    LIABILITY

-------
volumes are calculated or estimated. A detailed characterization
of  the environmental  setting,  including  hydrogeological
characteristics and the nature of  surrounding  populations, is
prepared.
  The focus of the report is environmental impairment (Fig. 1).
The risk assessment procedure takes into account four factors:

• Environmental Routes — the pathways, such as surface water,
  groundwater and air, by which materials could move off the
  premises of the facility into the environment
• Target Populations — the local human, plant and animal popula-
  tions that could be harmed by any materials moving off-site
• Facility Operations and Practices — the design, layout, opera-
  tions and practices of the former site
• Characteristics of Materials — the generic hazards of those
  materials that could migrate and  thus  affect off-site  target
  populations

  The factors are integrated to arrive at a full understanding of
the potential for on-site and off-site impairment as well as the pos-
sible consequences of the identified exposures. The assessment thus
affords a qualitative estimation of the actual and potential liabilities
facing a company.

INTRODUCTION

SITE  HISTORY

Site Description and Sources of Information
History
Technology and Site Layout

ENVIRONMENTAL SETTING

Geology and Soils
Hydrogeology
Surface Water

RISK ASSESSMENT

Environmental Exposures
Current Conditions
Target Populations

CONCLUSIONS AND RECOMMENDATIONS

                           Figure  1
                   Historical Risk Assessment
                         Report Format
  The management of uncertainty is by necessity a major compo-
nent of the analysis. The significant data gaps must be identified
and also evaluated in terms of their potential consequences and
influence on corporate liabilities. Informational needs can then be
prioritized and an appropriate response involving further research
or field investigations can be implemented to reduce or eliminate
crucial uncertainties.
  Historical risk assessment is by nature  an art, requiring the
investigator to know exactly what to look for, what is important
and unimportant, and what might present an environmental liability
in future years.

SITE REMEDIAL INVESTIGATION
  The risk assessment goes beyond discovering and evaluating
environmental risks by providing the creative insight needed to
develop programs aimed at reducing these risks. Based on the
historical  assessment,  recommendations to address identified
liabilities or remaining areas of uncertainty through environmen-
tal monitoring programs. Some combination  of test pits, soil
borings and groundwater monitoring wells typically is appropriate
to determine if residues remain, to establish the location and extent
of the residues, to determine if off-site migration has occurred and
to characterize the areal and vertical extent of contamination.
  In the absence of the constraints created by an imposed cleanup,
the remedial investigation can be carefully designed to obtain the
necessary information in a cost-effective fashion. A well-executed
site remedial investigation quantifies the various risks identified
or postulated in the assessment in terms of the predominant risk
factors.
  Conclusions of the investigation are drawn from an evaluation
of the significance of these quantitative and qualitative findings
as they relate to the identified environmental hazards. If remedial
actions  are warranted, the remedial  investigation guides any
feasibility studies required to plan and implement the necessary
actions.

CONCLUSION
  In conclusion, property owners and investors may be faced with
a complex range of liabilities arising from potential environmen-
tal impairment created by former industrial activities at a site. While
the nature of the potential contaminants may be predictable, a host
of variable factors affecting impairment and liability are associated
with the history and environment of the  site.  Historical risk
assessments of any former industries are prudent risk management
tools for companies  attempting to maintain a pro-active stance in
today's regulatory climate. They are absolutely necessary to achieve
a clear understanding of potential environmental liabilities and
make informed decisions on the appropriate responses.
  Risk  assessments  and  followup site  remedial investigations
initiated by a company can pre-empt potential enforcement actions
if carefully designed, implemented and  documented. Liabilities
often can be more effectively managed and remedial costs usually
reduced when companies are given a free hand in remedial plann-
ing.  Historical risk assessment provides an invaluable foundation
for an effective strategy of response based  on the reconstruction
of historical knowledge and the implementation of cost-effective
risk  management solutions.
                                                                                                              LIABILITY     47

-------
                                   Corporate Successor  Liability
                                       For  Environmental  Torts

                                              Mell J. Branch Roy, Esq.
                                          Jacobs Engineering Group, Inc.
                                                 Lakewood, Colorado
ABSTRACT
  Liability for corporate successors has become more prevalent
as an issue for courts to address, specifically regarding environ-
mental torts. In many state courts, corporate successor liability
raised by environmental issues is a case of first impression. The
precedent set forth in products liability cases has been relied upon
to encompass environmental tort cases. Public policy considera-
tions compelled a minority of courts to construct products liability
principles by extrapolating or modifying corporate tort and con-
tract law principles.

INTRODUCTION
  Corporate acquisitions involving the purchase or takeover of
companies have increased in recent time. A major factor to be con-
sidered in evaluating a proposed corporate acquisition, merger or
consolidation is the incurrence of debts or liabilities  by the pur-
chasing corporation. Following traditional principles of corporate
law, the structure of the acquisition is the premise for  successor
liability. The structure of the acquisition has  lost impact on the
imposition of corporate  successor liability for environmental torts.
  The U.S. EPA has confronted the issue of corporate  successor
liability in  the identification of potentially responsible  parties
(PRPs) pursuant to the Comprehensive Environmental Response,
Compensation and Liability Act (CERCLA),  42 U.S.C. 9601 et
seq. as amended by the Superfund Amendments and Reauthori-
zation Act (SARA) of 1986. Arguably, CERCLA liability may be
imposed  on a successor corporation where  the predecessor cor-
poration was an owner or operator of a hazardous waste manage-
ment facility or a generator or transporter of hazardous wastes,
as defined by CERCLA Section 107.
  Based on the interpretation of several theories, various factors
must be considered in determining the imposition of liability on
successor corporations.  Such determinations are important to the
U.S. EPA  in the identification  of PRPs, to  corporations con-
sidering the acquisition of a company involved with hazardous sub-
stances  and  to corporations  which previously acquired  such
companies.
  Since the  rules in corporate law vary from  state to state, this
paper reviews the general theories and principles of corporate suc-
cessor liability. Included in the paper is a discussion of the prece-
dent set forth by those courts addressing liability issues raised by
environmental claims.

TRADITIONAL RULE OF NON-LIABILITY
FOR SUCCESSOR CORPORATIONS
  The traditional rule of corporate law for successor  liability sets
forth that a corporation acquiring all, or part, of the assets of
                                                          another corporation does not assume the liabilities and debts of
                                                          the predecessor.1 If one corporation sells all its assets to another
                                                          corporation, the purchasing corporation generally is exempt from
                                                          the selling corporation's debts and liabilities.2 The general rule of
                                                          nonliability for the purchasing corporation in an assets transfer
                                                          is subject to four well-established exceptions. In the instance of
                                                          an assets transfer, liability may attach:

                                                          • where there is  an  implied or express assumption of liabilities
                                                          • where the transaction  amounts to  a  consolidation or merger
                                                          • where the successor is a mere continuation of the predecessor
                                                          • where the transaction was fraudulent, not made in good faith
                                                            or made without sufficient consideration3
                                                          Exceptions to the traditional corporate law rule were premised on
                                                          corporate and commercial law  principles. The  exceptions were
                                                          developed to protect the  rights of minority shareholders, to pro-
                                                          tect the interest of creditors by providing for the claims of known
                                                          creditors in transactions between purchaser and  seller, as well as
                                                          to determine liability in tax assessment cases.4
                                                            The traditional corporate law rule was not constructed to apply
                                                          to tort actions. The  exceptions to successor non-liability do  not
                                                          impact the degree of incompatibility  between corporate law  and
                                                          tort law principles of liability.5  In tort cases such as products
                                                          liability, application of the rule produced harsh and unjust results.
                                                          Consumers injured  by defective  products  were left with little
                                                          recourse where corporate ownership of the manufacturer was trans-
                                                          ferred.6 In Kloberdanz v. Joy Manufacturing Co., 288 F.  Supp.
                                                          817 (D. Colo. 1968),  the transaction between the two corporations
                                                          involved the purchase of certain assets and certain liabilities  of a
                                                          manufacturer of oil drilling equipment. The transaction was not
                                                          held to be a merger  or a consolidation of the two corporations.
                                                          The transactions appeared to be a legitimate sale conducted at arms
                                                          length and there was no indication of fraud. The purchasing  cor-
                                                          poration was found not to be be a continuation of the selling  cor-
                                                          poration, but a separate entity. Successor liability was not imposed
                                                          by the Court.
                                                            Such results were rebuked as being inconsistent with public policy
                                                          considerations.7 To  provide an avenue of relief for product lia-
                                                          bility claimants, some courts abandoned  the traditional rule of  non-
                                                          liability and expanded its exceptions.8 The following section dis-
                                                          cusses the exceptions to the rule of successor non-liability and the
                                                          impact of strict liability torts.
                                                          Exceptions to the Traditional Rule
                                                          Express Assumption Exception
                                                            Based on the terms of the purchase agreement, a corporation
                                                          which acquires the assets of another corporation may also incur
48
LIABILITY

-------
successor liability. The purchasing corporation may incur liability
by an express or implied agreement to assume such liability under
the terms of the purchase agreement.9
  Where the language is broad (e.g., company assumes all debts,
obligations and liabilities of X company), the purchasing corpo-
ration assumes the risk of future claims and unknown or contin-
gent liabilities, even though the specific agreement does not refer
to claims arising from future events.10 In the environmental tort
case of Philadelphia Electric Company v. Hercules, Inc., the court
found that Hercules broadly assumed all liabilities incurred by the
predecessor company as of closing date, subject to a few limited
exceptions. Judge Higginbotham wrote "... it is of no consequence
that the specific liability at issue is not enumerated... Unless this
liability conies within one of the express exceptions, Hercules may
be held to have assumed it.1
  The effect of a disclaimer of any non-disclosed liability by the
successor in a purchase agreement may avoid a finding of an express
or implied assumption of future tort claims. To have effect, the
language  limiting   liability  must  be  clear, specific  and
unambiguous2.
  Language in the purchase agreement limiting or disclaiming lia-
bility of the purchasing corporation merely binds the  parties to
the agreement. The buyer and seller corporations may regulate how
liability for claims will be allocated  among  themselves, but such
provisions have no  effect on the rights  of third parties.13. In
Mardan Corp. v. C.G.C. Music, Ltd., 804 F.2d 1454 (9th Cir.
1986), a case  raising liability issues  under CERCLA,  the Court
found that contractual agreements apportioning CERCLA liability
between "responsible parties"  will not prejudice the right of
government to recover closure or cleanup costs  from any respon-
sible party.

Merger Exception
   If-a corporation is acquired by merger, consolidation or sale of
stock, the transferee will become liable for claims against the
acquired company.14 Pursuant to procedures prescribed by state
statute, the incident of merger is conspicuous and the assumption
of liability by the successor corporation is  clearly defined.
  A merger results from the absorption of a corporation and the
practically contemporaneous dissolution of the acquired corpora-
tion as  a legal entity.15 Although the terms merger and consoli-
dation are used interchangeably, the transactions are distinct. In
a consolidation, existing corporations dissolve and form  a new cor-
porate entity which succeeds to all properties, powers, privileges
and liabilities of the predecessors.6
  A New Jersey case raising issues of environmental tort is illus-
trative of successor liability and statutory merger. The New Jersey
Department of Environmental Protection v. Ventron,1'1 involved
a claim for damages resulting from contamination by a chemical
company's mercury processing plant. Wood  Ridge Chemical Cor-
poration (Woodridge) was a subsidiary of Velsicol Corporation
(Velsicol). The parent corporation created the subsidiary to acquire
the assets of the F.W. Berk & Co., Inc. and to continue operating
the  mercury processing plant. Velsicol sold 100% of the Wood
Ridge stock to Ventron Corporation  (Ventron);  subsequently, the
mercury processing plant operations ceased and the property was
sold. The court imposed successor liability on Ventron  as a result
of its merger  with Wood Ridge.18
If,  for all practical purposes, a sale  of assets effects a  merger or
consolidation, the transaction may afford imposition  of succes-
sor liability on the purchasing corporation.19

DEFACTO MERGER DOCTRINE
  By applying the  de facto  merger doctrine, some courts have
extended successor liability to the corporate law rule. A de facto
merger is a judicially created concept to protect: (1) minority share-
holders; (2) rights of creditors;  and (3) claims  of tort victims.20
One distinction between an assets transfer and a merger is that a
merger  encompasses certain rights and liabilities which generally
are not included in a sale of assets where the parties are unrelated
and adequate consideration is present.
  In Shannon v. SamuelLangston Co., 379 F. Supp. 797 (W.D.
Mich 1974), a products liability case, the court imposed liability
and found de facto merger as distinguished from an ordinary sale
of assets. The precedent set forth in Shannon provides a de facto
merger may be found based on the following criteria: (1) there is
a continuation of the enterprise of the seller corporation so that
there is a continuity of management, personnel, physical location
assets and general business operations; (2) there is a continuity of
shareholders which results from the purchasing corporation paying
for the acquired assets with shares of its own stock ultimately
coming to be held by the shareholders of the seller corporation
so that they become a constituent part of the purchasing corpora-
tion; (3) the seller corporation ceases its ordinary business opera-
tions, liquidates, and dissolves as  soon as legally possible; and
(4) the purchasing corporation assumes those liabilities  and obli-
gations of the seller ordinarily necessary for the uninterrupted con-
tinuation  of  normal business  operations of  the  seller
corporation.21
  The de facto merger doctrine developed principles which allowed
tort victims to recover in light of a corporate change of owner-
ship. In Knapp v. American Rockwell Corp.22, a products liability
case involving a transfer of assets for stock of the acquiring com-
pany, the court found a de facto merger where the asset transfer
merely resembled a merger. The court presented specific factors
to consider in a de facto merger case: (1) factors establishing in-
substantiality of continued existence; (2) brevity of the continuance;
(3) contractual requirements  for dissolution; (4) prohibition of
further normal business operations; and (5) the character of the
predecessors remaining assets.23
  To illustrate a consistent holding for an environmental tort
action, liability of a corporate successor was imposed in  Philadel-
phia Electric Company v.  Hercules, Inc24 where the predecessor
firm allegedly caused hazardous waste contamination. In applying
Pennsylvania law,  District Court Judge McGlynn found the suc-
cessor corporation liable on the basis  that the transfer  of assets
between the successor and predecessor companies constituted a de
facto merger.
  A key factor to buttress a finding of de facto merger for successor
liability is a transfer of stock as consideration.25 Where  the assets
are sold  for cash, the relationship of the stockholders to their
respective corporations remains unchanged, and there is an absence
of continuity of shareholders, a finding of de facto merger is in-
consequent.26
   The basis for finding a de facto merger has similarity to an appli-
cation of the continuation exception. There are, however, separate
grounds for which liability is imposed on the successor corpora-
tion. Although the two exceptions  are related, a de facto merger
generally entails a stock-for-assets acquisition that is not required
for the continuation exception.27 Evidence that a de facto merger
occurred often is sufficient to impose liability under the  continua-
tion theory.28

CONTINUATION EXCEPTION
Mere Continuation Theory
   Successor liability using the "mere continuation" theory is based
on the premise that, despite the change in name or form, the suc-
cessor corporation has virtually the same identity as the  predeces-
sor. The central theme supporting the "mere continuation" theory
is the emphasis on the continuation of the corporate entity rather
than the continuation of business  operations.29 The critical fac-
tors in applying the traditional "mere continuation"  exception
include the following:

•  continuity of shareholders; directors; officers; and business oper-
   ations.30 To find that a successor is a mere continuation of a
   predecessor, the court may consider whether all or a  large por-
   tion of the assets of the predecessor  were purchased by the
                                                                                                             LIABILITY    49

-------
  successor or if the predecessor goes out of business at the time
  of the sale of its assets or shortly thereafter.31

  Changes in the development of product liability principles have
allowed some jurisdictions to expand the mere continuation theory
to be more applicable to strict  liability torts. The continuity of
enterprise emphasizes the continuation of the entire business oper-
ation rather  than  the mere continuity of  the corporate entity
represented by its  officers directors and shareholders.32

Continuity of Enterprise Theory
  Continuity of enterprise is indicative of the courts attempt to
address the competing  issues between corporate and tort law.
Driven by public policy  considerations to provide some recourse
to the injured consumer in products liability cases, some juris-
dictions have established precedent which facilitates the imposi-
tion of successor liability.
  In Cyr v. B. Offen & Co., the  guidelines are initially set  for con-
tinuity of enterprise in establishing successor liability." The court
expanded the (mere) continuation exception  by  eliminating  the
common ownership requirement and relied on strict liability policy
rather than corporate law.33 Although there  was an absence of
continuity of ownership between  the predecessor and successor,
the court found such requirements unimportant where the manu-
facturing entity itself continued in terms of the physical plant,
employees, supervisors  and product.34
  The ruling in Cyr provided that the policy reasons supporting
strict liability also support successor liability.35 The Cyr court
found "the very existence of strict liability for manufacturers  im-
plies a basic judgment that the hazards of predicting and insuring
for risk from defective products are better borne by the manufac-
turer than by the consumer.36 Accordingly, the successor is in a
much better position, due to  experience and expertise, to gauge
the risks and the costs of meeting them.37 Since the successor  will
reap the benefits of the  predecessor's accumulated good will, the
Cyr finding provides the  successor should also bear the risks despite
the fact that it was not the legal entity that placed the product in
the stream of commerce.
  A Michigan Supreme Court products liability case, Turner v.
Bituminous Casualty Company, rejected traditional corporate
criteria for imposing successor liability for defective products by
ruling that tort, and not corporate law principles were  applica-
ble.38 The court's  rationale provided that imposing  liability for
defective product injuries should not be premised on whether the
acquisition was affected through  a transfer of stock or purchase
for cash.39 Focusing on  continuity of enterprise rather than con-
tinuity of ownership, the Michigan court set forth that  a prima
facie case for successor products liability could be found when the
following facts are  present: retention of key personnel; same assets;
same general business operations; same corporate name, succes-
sor holds out to the public as being the same as the predecessor;
successor corporation assumes liabilities and obligations  incident
to continuation of normal business operations; and the predeces-
sor liquidated and dissolved after distributing cash consideration
received from successor.40 Although the Turner decision  deviates
from the corporate law  rule, the ruling is congruent with expan-
sion of the continuation exception.41

Product Line Exception
  In stating its policy rationale for the decision in the New Jersey
Department of Transportation v.  RSC Resources, Inc., the court
explained that the rationale of strict enterprise liability was apropos
to environmental tort liability.42 In Diamond Head Oil Refining
Company (Diamond Head), the predecessor corporation discharged
wastes into a nearby lake between 1946 and 1973. PSC Resources,
Inc. (PSC), the successor corporation, purchased all of the stock
of Diamond Head, which ultimately dissolved. Subsequently, PSC
acquired all of Diamond Head's assets.  According to the facts
presented in the opinion, "PSC was incorporated for the purpose
of acquiring the stock and/or assets of Diamond Head," supra,
at 1153. PSC acquired the Diamond Head assets for less than '100.
The successor corporation,  PSC, continued the general line of
business as the predecessor corporation. The Court held that a cor-
poration which purchases the assets of another corporation and
continues its general (product) line of business is strictly liable for
the environmental torts of its predecessor.
  In applying the productline exception, the courts may consider
the following factors:  whether the successor deliberately exploited
the predecessor's  established reputation  and good will  as  a
manufacturer of the product line at issue;  whether  the products
manufactured by the successor are similar or identical to those
manufactured by the predecessor; and, the successor's use of the
predecessor's physical plant, equipment and continuity of corporate
personnel.43
  General applicability of the "mere continuation,"  continuity of
enterprise or product line exceptions to noncorporate business
entities has not been established. In Tift v. Forage King Industries.
Inc44, the issue was whether an incorporated sole proprietorship
is responsible for liabilities of the sole proprietorship. The Court
applied the corporate law rule and imposed liability  based on the
continuity of enterprise.45  As opined by the Court, there was sub-
stantial identity between predecessor and successor, even though
the name organizational  form was changed.46
  Successor liability was imposed on a state created entity in U.S.
v. Metropolitan District Commission." Citing Cyr, the court held
the successor state agency liable  for violations by its predecessor
of the Clean Water Act,  33 U.S.C. Section 125 (et seq.4*.

Fraud Exception
  A fraudulent conveyance may impose successor liability where
the parties to the transaction do not exercise good faith.49 Liabil-
ity may attach if the parties fail to prove that the transaction was
made in  good faith and  for value.

STRICT LIABILITY AND ENVIRONMENTAL TORTS
  Exceptions to the traditional successor non-liability rule and sub-
sequent modifications were the results of public policy considera-
tions to provide injured parties some redress despite a change in
corporate ownership. A conspicuous constituent in both products
liability and environmental ton claims is that of strict liability.50
Pollution cases are similar to products  liability cases.  As stated
by  the Court in PSC, "The salient characteristic in product lia-
bility cases in which successor liability was found is that each tort
was declared to be one of strict liability.51"
  Rules of strict liability for defective products and environmen-
tal  causes of action have analogous public  policy underpinnings.
The impetus behind recasting corporate law principles was public
policy considerations such that  successor  liability  may be
imposed.'2 Policy considerations include effective cost-spreading,
risk-avoidance devices and availability of a defendant from whom
an  injured party may seek redress. In PSC and Hercules the court
found policy rationale in environmental torts analogous to that
of  products liability.
   Based on legislative history and developing case law, it  is estab-
lished that CERCLA imposes strict liability and may be applied
retroactively. Some state  environmental statutes, like CERCLA,
also impose strict liability and are applied retroactively.51 As such,
strict liability for hazardous waste and environmental torts applies
to successor and predecessors, whether or not the parties were aware
of  the hazard.

RISK AVOIDANCE IN ACQUISITIONS
  Protections once afforded corporations  purchasing  assets are
minimal.  It  is advisable that the purchaser become "an informed
buyer."  With the  cost of cleanup and removal at  astronomical
levels, a  buyer should beware of latent environmentaJ hazards.
  Whenever a corporation is seeking to acquire another corpora-
tion, the assumption of unknown or  contingent liabilities should
factor into the purchasing decision.
50    LIABILITY

-------
  The following are a few suggestions for potential purchasing cor-
porations; and depending on the acquisition, some suggestions are
more practicable than others. A buyer should obtain as much avail-
able information  as possible to make an accurate assessment of
the environmental risks. An "environmental" audit of the seller's
assets, operations and known  liabilities may be conducted. The
buyer  should investigate past and present land uses and business
operations. Whether hazardous waste was generated or stored by
the seller should be  determined. If the seller's operations entailed
use of hazardous  materials, the buyer should ascertain the extent
of such use. (In this instance, the buyer may request a history of
the seller's waste disposal practices, including on-site and off-site
facilities.) An inspection of corporate and government records to
determine regulatory compliance history may alert the buyer to
any prior violations or regulatory non-compliance.
  After a thorough investigation of the seller's operations, sale
negotiations should err on the site of caution. The purchase agree-
ment should be clear and unambiguous as to  the assumption or
disclaimer of liabilities. The parties may enter an indemnification
agreement to indemnify the seller should an action arise from the
predecessor's activities. The indemnification should clearly indem-
nify a party from strict liability actions  such as environmental torts.
The buyer may insist the seller obtain liability insurance for dis-
charge or release of hazardous substances. The seller should be
aware that the insurance industry is becoming more reserved in
writing insurance policies for parties that may have environmen-
tal hazard claims.

CONCLUSION
  The courts modified and recasted successor non-liability princi-
ples to accommodate tort victims. In view of public policy con-
cerns, the courts  facilitated the imposition of successor liability
by  expanding the exceptions to the traditional rules.  As acquisi-
tions of corporations become more frequent, acquiring corpora-
tions may require more information on the seller's business prior
to purchase. Since the structure of the acquisition is no longer the
focal point  of successor liability,  corporations  must  take pre-
cautions to avoid future tort claims resulting from past predeces-
sor activities.

REFERENCES
  1. See generally 19 Am Jur 2d Corporations Sections  1546-1553; 15 W.
    Fletcher, Cyclopedia of the Law of Private Corporations Section 7122
    (Rev. perm.  ed. 1983), [hereinafter Fletcher].
  2. Aylward, Robert  E. and Aylward, Janice S., Successor Liability for
    Defective Product-Misplaced Responsibility.  13 Stetson Law Review,
    555 (1984). (citations omitted).
  3. Kloberdanz v. Joy Mfg. Co., 188 F. Supp. 817 (D. Colo.  1968); Knapp
    v. North American Rockwell Corp., 506 F. 2d 361 (3d Cir. 1974). Amer.
    Law Prod. Liab., 3d Section  7:1 (1987).
  4. New Jersey Transportation Department v. PSC Resources, Inc., 175
    N.J. Super 447;456;419 A. 2d 1151, 1956 (1980). Comment, A Restora-
    tion of Certainty: Strict Products Liability and Successor Corpora-
    tions., 43, Ohio St. Law Jour. 441,442-43 (1982).
  5. PSC Resources, 175  N.J.  Super, at 456, 419 A. 2d at 1156.
  6. McKee v. Harris—Seybold Co., 109 N.J. Super. 555 (Law Div. 1970),
    aff'd on  other grounds, 118 N.J. Super 480 (App. Div. 1972); See
    Kloberdanz, note 3.
  7. See Comment, supra note 6.  19 Amer Jur 2d Corporations Section
    1554; Annotations: Successor Products Liability: Form of Business
    Organization of  Successor or Predecessor  as Affecting Successor
    Liability. 32 ALR 4th 196.
  8. See Cyr v. B. Offen & Co., 501 F.2d 1145 flst cir. 1974); Turner v.
    Bituminous Casualty Co., 397 Mich. 406, 244  N.W. 2d 873 (1976);
    Annotation: Products Liability of Successor Corporation for Injury
    or Damage Caused by Product Issued by Predecessor. 66 ALR 3rd 824.
  9. Hanlon v. Johns-Manville Sales Corp., 599 F. Supp. 376 (N.D. Iowa,
    1984) (applying Iowa law); Emrich v. Kroner, 79 App. Div. 2d 854,
    (4th Dept.) 434 NYS 2d 491 (1980). Annotation: Liability of Succes-
   sor Corporation for Injury or Damage Caused by Product Issued by
   Predecessor. 66 ALR 3d 824.
10. Id.; Philadelphia Electric Company v. Hercules, Inc., 762 F.2d 303,
   309 (3rd Cir. 1985).
11. 762F.2d at 309.
12. Id. at 310.; Mudgett v. Taxson Machine Co., 709 SW 2d 755 (Tex.
   App. Corpus Christi, 1986); Lopata v. Bemis Company, 383 F. Supp.
   342 (E.D.  Pa. 1974), vacated, 517 F. 2d 119B (3d Cii.), judgment re-
   instated, 406 F. Supp.  521 (1975), aff'd 546 F.2d 417 (1976); Adams
   v. General Dvnamics Corp., 405 F. Supp. 1020 (N.D. Cal., 1975)
   (applying California law); Am Law Prod Liab 3d Section 7:4 (1987).
13. Grant-Howard Associates v. General Housewares Corp., 63 NY 2d
   291, 482 NYS 2d 225, 472 NE 2d 1, (1984). Kadens, Michael G., Prac-
   titioner's Guide to Treatment of Seller's Product Liabilities in Assets
   Acquisitions, 1O Univ. of Toledo. Law Rev., 1 (1978).
14. See Fletcher, supra note 1, sections 7041, 7122, 7205; Jackson v. Dia-
   mond T. Trucking Co., 100 N.J. Super. 186 (Law Div. 1968), Anno-
   tation: Products  Liability of Successor Corporation for Injury or
   Damage Caused by Product Issued by Predecessor 66 ALR 3d 824.
15. See Fletcher, supra Section 7041; Shannon v. Samuel Langston Co.,
   379 F. Supp. 797  (W.D. Mich. 1974); Anders for Anders v. Jackson-
   ville Electric Authoritv, 443 So. 2d 330 (Fla.  App. 1983).
16. Fletcher, supra note 1.
17. Department of Environmental Protection v. Ventron Corporation 94
   N.J. 473,  468 A.2d 893 (1983).
18. Id. at 463.
19. See Fletcher, supra note 1, Section 7205; Bererard v. Kee Manufac-
   turing Co., 409 So. 2d 1047 (1982).
20. Farris v. GlenAlden Corporation. 393 Pa. 427, 143 A. 2d 25 (1958).
   Fletcher, supra note 1, sections 7122-23. See Kadens, supra note 16;
   Defacto Merger of Two Corporations. 20 Am Jur POP 2d 609.
21. Supra note 19, 379 F. Supp at 801.
22. 506 F. Supp 361  (3d Cir.  1975)
23. 506 F. Supp. at 367-369.
24. 587 F. Supp 144 (E.D. Pa. 1984), Reversed on other grounds, 762 F.
   2d 303 (3rd Cir.  1985).
25. Dayton v. Teck Stow & Wilcox Co., 739 F.2d 690 (1984); Ray v. Alad
   Corp., 19 Cal 3d 22, 136  Cal Rptr. 574, 560  P.2d 3 (1977); McKee
   v. Harris-Seybold,  109 N.J. Super 555, 264 A.2d 98 (Law Div. 1970)
   aff'd 118 N.J. Super. 480, 288 A.2d 585 (App. Div.  1972) Am Law
   Prod. Liab. 3d Section 7:7 (1987)
26. Am. Law Prod.  Liab. 3d Section 7:6 (1987).
27. In Ray  v.  Alad, the court indicated that a purchase  of assets is not
   necessary  to finding a de  facto merger 19 Cal. 3d 22, 28 (1977).
28. De facto Merger of Two Corporations, 20 Am Jur POP 2d 609; An
   notation:  Similarity of Ownership or Control as Basis for Charging
   Corporation Acquiring Assets of Another \vith Liability for Former
   Owner's Debts. 49 ALR 3d 881. Kadens, supra note 16.
29. Annotation: Products Liability: Continuation of Business Enterprise
   or Product Line  by Successor Corporation. 46 Am Jur POP 313.
30. Id., Travis v. Harris Corporation, 565 F.2d 443, 447 (7th Cir. 1977),
   Flecther, supra note 1, Section 7122.
31.  See supra note 35, Annotation 66 ALR  3d 824.
32.  See Cyr v. B. Offen & Co., 501 F.2d 1145 (1st  Cir.  1974) (applying
   New Hampshire  law).
33. Id. at 1151-53.
34. Id. at 1151
35.  Id. at 1154. See Comment, Continued Expansion  of Corporate Suc-
    cessor Liability in the Products Liability Arena, 58 Chicago Kent Law
    Rev. 1117 (1980).
36.  Id.
37. Id.
38.  397 Mich. 406, 244 N. W. 2d 873 (1976).

39. Id. at 423, 244 N.W.  2d  at 880.

40. Id. at 430, 244 N.W.  2d at 883-884.
                                                                                                                     LIABILITY    51

-------
 41 Id.                                                                  49. Annotation: Similarity of Ownership or control as Basis for Charging
 42  Suora note at 1161                                                     Corporation Acquiring Assets of Another with Liability for Former
       ^                                  ,„,„   ,„„  ffn   „.           Owner's Debts 49 ALR 3d 881. See. 66 ALR 3d Section 4d.
 43.  Ray v. Alad Corporation, 19 Cal 3d 22, 136 Cal Rptr 574, 560 p. 2d
     3 (1977). 63 Am Jur 2d, Products Liability Section 175. 46 Am Jur        50- n(~- W"-
     POP 2d 313.                                                        51. PSC, supra at 1159.
 44.  108 Wis 2d 72, 322 NW 2d 14.                                        52. Wilson v. Fare Well Corp., 140 N.J. Supr. 476 (Law Div. 1976). 66
       o ,„• ,_, ,,, v, „, ,j    ,,.,,, A, „ ,. ,. ,,,                         ALR 3d 824. Corporations Under CERCLA. 9 Chem & Rad Waste
 45.  108 Wis 2d, 322 N.W. 2d at 16-17 32 ALR 4th 172.                        Li(  Rptr  , J3 (19g4)

 46.  Id. See, Cyr supra.                                                  53. Predergast, Winifred and Shanahan, Robert, "Successor Corporate
 tn  ซ unr iซi /n  Mace  ios<>                                            Liability for Hazardous Waste Cleanup and Removal Under the Penn-
 *Tป*  ฃ-J CIx\^ iJJl llj*. IVlOdd. 17OJ1.                                             .    ,  -^i    —       _       , .  - ,   -     M ... .   ,.  ,_.
                                                                           sylvania Clean Streams Law and the New Jersey Spill Act.  ' 4 Temple
 48.  Id. at 1361.                                                            Envl'l Law and Tech.  Int., 61  (1985).
52     LIABILITY

-------
                                  State  and  Local Jurisdiction At
                     Federal  Facilities  with  Hazardous  Waste Sites

                                              John ฃ. Cromwell, Ph.D.
                                                  Alyson A. Hennelly
                                                  Roy F.  Weston, Inc.
INTRODUCTION
  Federal facilities no longer enjoy a presumption of immunity
with regard to Federal, state and local environmental laws. This
situation has translated into an increased number of statutory areas
where state and local authorities now have jurisdiction. Amend-
ments to major environmental statutes and Executive Orders have
placed constraints on when and where Federal facilities can exer-
cise immunity—that protection from laws reserved to the nation
itself and  referred to as sovereign immunity. Congress also has
enacted into law, through major program amendments to RCRA
and SARA, waivers of sovereign immunity for Federal facilities.
This means, in certain situations, that  Federal facilities  not only
have to comply with Federal, state and local environmental laws,
but also must obtain applicable state and local permits. State and
local authorities do have jurisdiction over Federal facilities.  The
number of instances where Federal facilities are entitled to sovereign
immunity  is decreasing and will continue  to decrease.
  Executive Order  12088,  signed October  1978, requires Federal
facilities to attempt to comply substantively with all Federal, state
and local environmental laws. Heads of each executive agency are
responsible for insuring that all necessary actions are taken for
prevention, control and abatement of environmental pollution with
respect to  Federal facilities as well as activities under control of
the agency. The executive Order also requires agencies to submit
to the Office of Management and Budget (OMB) an annual plan
for the control of environmental pollution. Since  a number of
environmental statutes allow the federal government to delegate
at least part of the federal  program to the states to operate (after
authorization), this has meant that Federal authorities have juris-
diction as  well as state and local authorities.
  Because of the nature of the RCRA program,  where states
become  authorized to operate the RCRA program in lieu of the
Federal  authorities,  states  are now  controlling  those areas
previously only in the purview of the Federal government.
  However, no Executive Order requires a Federal facility to obtain
permits from  state and local agencies unless  specifically directed
to in an environmental statute. This directive to comply is referred
to as a waiver—it is a waiver of the sovereign's immunity from
the laws of constituent governments. This occurs when the federal
facility's normal sovereign  immunity has been specifically waived.
  Although Congress recently has placed limits on the immunity
of Federal facilities from compliance with major environmental
laws, Federal  facilities still can be exempted from compliance for
reasons  of national security. Without  this  exemption, Federal
facilities must comply with Federal, state, interstate and local laws
that are consistent with the waivers in major environmental laws
such as the Clean Air Act, Clean Water Act, RCRA, SARA and
Safe Drinking Water Act.

RCRA and Superfund Authorities
  Examining two specific statutes and how they affect Federal
facilities will demonstrate the trend of decreasing immunity for
Federal facilities.

RCRA and Mixed Waste
  When the U.S. EPA first asserted that mixed waste (which is
hazardous waste mixed in any part with radioactive waste and
typically found at DOE weapons facilities), was subject to the pro-
visions of RCRA, the DOE took the position that RCRA was
inconsistent with the Atomic Energy Act and thus was  not
applicable. After an adverse court decision regarding DOE's Oak
Ridge facility (the LEAF decision), DOE changed its position,
saying that RCRA  applied  only to purely hazardous, non-
radioactive  wastes.
  After much discussion between the DOE, the U.S. EPA and the
NRC, DOE published a policy that gave U.S. EPA control, through
RCRA, over the hazardous component of radioactive mixed waste,
but still reserved total control over both the hazardous and radio-
active components of radioactive mixed waste with a specific
activity over 100 nanocuries/g of waste. DOE eventually changed
this position in  June 1987, to give U.S.  EPA control over the
hazardous component of all mixed waste, regardless of the radio-
activity. Thus, DOE weapons facilities are now subject to the
requirements of RCRA as they apply to mixed waste.

Superfund
  Section 120 of SARA requires that Federal facilities be treated
the same as non-governmental entities in complying with the pro-
visions of SARA, including being assessed and placed on the NPL,
if appropriate.
  The U.S. EPA had determined that the most of  the Federal
facilities that could be placed on the NPL currently have RCRA
operating units within the facility property boundary. Applying
U.S. EPA's non-federal facilities RCRA policy to Federal facilities
would result in placing very few Federal facilities on the NPL. The
U.S. EPA felt that this result—placing few Federal  facilities on
the NPL—was inconsistent with the intent of SARA. Therefore,
the U.S. EPA proposed a revised policy for placing Federal facilities
on the NPL. The new policy would allow placing on the NPL
Federal facility sites  that  may be subject to the corrective-action
authorities under RCRA. Normally, NPL listing is deferred in this
                                                                                            POLICY ASSESSMENT    53

-------
case unless one of the following criteria is met:

• The owner/operator is bankrupt
• The owner/operator has lost RCRA interim status and there are
  additional indications that the owner/operator will be unwill-
  ing to undertake corrective action
• The owner/operator has shown an unwillingness to undertake
  corrective action at a site (based on case-by-case analysis)

  At this time, the U.S. EPA also is reviewing a March 1986 facility
boundary policy, subjecting Federal facilities to RCRA corrective
action requirements and stating that a Federal  facility boundary
"is equivalent to the property-wide definition of facility at privately
owned or operated facilities." This is perhaps a future additional
limitation on the immunity of Federal facilities.
  On July 21, 1987, the U.S. EPA added 32 Federal facility sites
to the NPL. The U.S. EPA also reproposed adding seven Federal
facilities to the list.

The U.S. EPA Federal Facilities  Task Force
  Some states are still having a difficult time negotiating trie Federal
facilities in their state into compliance. The state of Washington
will no longer negotiate with the DOE and will only litigate to force
compliance.
  The U.S. EPA is having its own difficulties with enforcing com-
pliance at Federal facilities. After Congressional oversight hearings
in April 1987, where the U.S. EPA was charged with lax enforce-
ment at Federal  facilities by Congress, the U.S. EPA responded
by undertaking a two-part initiative:

•  Part One: Reviewing a  10-year old Memorandum of  Under-
  standing (MOU) between Justice and the U.S. EPA to see if the
  U.S. EPA can find room and  authorities to sue Government
  Owned Contractor Operated (GOCOs) and Contractor  Owned
  Contractor Operated (COCOs).
•  Part Two: Issuing a Federal facility compliance strategy. The
  proposed  strategy would allow the U.S. EPA to give Federal
   facilities Administrative orders to comply when negotiations
  break-down. The Administrative or  compliance  order would
  allow a state or citizen to file a citizen's suit under the provisions
  of environmental laws such as the Clean Water Act or RCRA
  if the Federal facility is not abiding by the order. In order to
  issue the order, the U.S. EPA needed to gain the consent of the
  agency running the Federal facility and the Department of Justice
  (DOJ).  DOJ rejected this plan.

  The DOJ  believed that it was unconstitutional for one agency
in the executive branch to sue another agency. The DOJ believed
that the President  as unitary executive, should resolve disputes
between the agencies.
   In response to these problems and tensions, the U.S. EPA has
formed a Task Force specifically directed to Federal facility com-
pliance. The Task Force is headed-up by Christopher Grundler and
is housed  in the U.S. EPA Office of Waste Programs Enforce-
ment (OWPE). The Task Force has three objectives:

•  Bring Federal facilities  into compliance
•  Develop the Federal facility hazardous waste docket as required
  by the 1986 SARA
• Develop policy and guidance on Federal facility issues, par-
  ticularly the RCRA-Superfund overlap
• According to the Task Force, their initial targets are the Idaho
  National  Engineering  Laboratory  (INEL),  Hanford   and
  Lawrence Livermore facility  in California.

MODEL AGREEMENTS
  Recent model agreements have established the U.S. EPA's right
to enforce hazardous waste laws under both the RCRA and Super-
fund laws.

Superfund - Twin  Cities
  The most recent agreement, the Twin Cities Agreement between
the U.S. EPA, the U.S. Army and the Minnesota Environmental
Pollution Control Agency sets a compliance and cleanup schedule
for the Twin  Cities Army Ammunition Plant.
  This is EPA's first agreement with the DOD regarding the DOD's
compliance with RCRA provisions at their facilities. The DOD has
estimated in a recent report that there are 3500 potential hazardous
waste sites at 529 military installations and has given estimates that
it may take over $10 billion to clean up the worst 400 to 800 sites
over the next  decade. The Twin Cities agreement is therefore the
first  of many.

Superfund - Idaho National  Engineering Laboratory (INEL)
  This agreement, reached July 1987, between the U.S. EPA and
DOE, is the first agreement under SARA. Both this agreement and
the Twin Cities agreement established the U.S. EPA administrator
as the final arbiter in disputes where the U.S. EPA and DOE cannot
agree.  The  agreements also allow citizens,  suits to  be the
enforcement mechanism. These two agreements—the Twin Cities
and the INEL agreements—settle the choice of law dilemma that
has faced the U.S. EPA—whether  to choose SARA Section 120,
or RCRA. The INEL agreement  is  a strong agreement because it
is a consent order and includes a clearly stated process and timetable
for resolving disputes  on  the  local and  regional level before
requiring resolution by the U.S. EPA administrator.

RCRA - Rocky Flats
  The Rocky Flats agreement was the first agreement between DOE
and  the U.S.  EPA regarding U.S.  EPA jurisdiction over mixed
hazardous and radioactive waste.  This agreement is considered
weaker than the INEL agreement  because, among other things,
the U.S. EPA cannot sue or fine DOE. There are no other enforce-
ment mechanisms in this agreement.

FEDERAL FACILITY CURRENT COMPLIANCE
  Federal facilities currently must comply with and obtain permits
under the following statutes:

• Clean Air - Air Permits
• Clean Water - Surface waster discharge permit and dredge and
  fill permit  (section 404 (bXO)
• Safe Drinking Water Act - Public drinking water supply system
  and Underground Injection Control (UIC)
• SARA - As specified by SARA, the U.S. EPA must establish
  a special Federal Agency Hazardous Waste Compliance Docket,
  open to the public. For each site in the docket, the U.S. EPA
  must ensure that a Preliminary Assessment (PA) is conducted
  by the facility within 18 months of enactment of the SARA.
  Within  30 months  of enactment. Federal facilities must be
  evaluated for  inclusion on the NPL where appropriate.

     Once listed  on the NPL, Federal facilities  will be required
     to initiate a RI/FS within 6 months of the listing and enter
     into an Interagency Agreement (1AG) with U.S. EPA within
     6 months of the U.S.  EPA review of the RI/FS. The U.S.
     EPA and the state must set and establish deadlines for the
     RI/FS. Facilities already on the NPL at the time of enactment
     must begin  such studies within one year.
     SARA also provides  for joint  U.S.  EPA/Federal Agency
     selection of the remedy, or final selection of the remedy by
     the EPA Administrator if the EPA and  the Federal agency
     are unable to reach agreement. Remedial action must begin
     within 15 months of the RI/FS completion. Public participa-
     tion and participation by state and local officials in this process
     is required. The remedial  action must  be completed  "as
     expeditiously as practicable." Agencies are required to submit
     a statement of the costs of cleaning up a site and the state-
     ment of the hazards to  human health and the environment
     in their annual budget submissions to Congress. Each Agency
     must submit an annual report  to Congress on the progress of
     the cleanup.
54    POLICY ASSESSMENT

-------
  - For those sites not on the NPL, State laws governing removal
    and remedial action will apply to Federal facilities, including
    enforcement.  States  may  also  apply more  stringent
    requirements to Federal facilities than are applied to non-
    federal facilities.
    The  President is authorized to  grant a national security
    exemption for Federal facilities. The President must notify
    Congress, and the exemption can be only one year in dura-
    tion, although the exemption can be renewed each year for
    a period not to  exceed 1 year.
• RCRA - Hazardous waste permits, by products (mixed waste).
• Floodplain/Wetland Law: evaluating permit application for
  projects affecting wetlands
• Wildlife Protection Statutes:


  - Migratory Bird Treaty Act
  - Fish and Wildlife Coordination Act
  - Bald and Golden Eagle Protection Act
  - Endangered Species Act
  - National Wildlife Refuge System Administration Act
  - The U.S. Fish and  Wildlife Service administers these Acts:
  - Cultural Resource Protection - National Historic Preservation
    Act and Historical and Archaeological Preservation Act. The
    Advisory Committee  on Historic Preservation has jurisdic-
    tion, as does the  State Historic Preservation Officer, and is
    involved in determining if significant cultural resources are
    present.
CONCLUSION

  Executive Order 12580, signed in January 1987, delegates to the
U.S. EPA responsibility for overseeing cleanups of sites listed on
the NPL. Responsibility for those sites not listed on the NPL and
removal actions other than emergencies remain the responsibility
of the heads of the executive agencies with jurisdiction, custody
or control over the site. There are continuing negotiations between
the U.S. EPA, DOJ, DOE and DOD over the extent to which
Federal facilities have to comply and what specific authorities U.S.
EPA has to enforce and oversee compliance.
  What this discussion of jurisdiction demonstrates is that it will
become increasingly  difficult for Federal  facilities to maintain
themselves  outside the  existing and  increasingly  structured
framework of Federal,  state,  interstate and local control over
hazardous waste.  Listing Federal facilities on the NPL will focus
more public attention on Federal non-compliance with environ-
mental laws. Because these agencies have to submit to Congress
budgets for funds to clean up hazardous waste sites on their
facilities, $10  billion submissions from DOD for  cleanup will
quickly focus more Congressional scrutiny on Federal facilities and
compliance.
  Although DOE and DOD facilities may be exempted this year
because of national security reasons, after publicity is focused on
those sites listed on the NPL and after the toxicological profiles
are published highlighting human health effects of chemicals found
at many of these Federal facility hazardous waste sites, it will be
harder to operate outside the framework of control. Eventually,
the public and Congress, through its oversight role, Will force DOD
and even those Federal facilities with national security exemptions
to comply with these environmental laws.
                                                                                                 POLICY ASSESSMENT     55

-------
                  Risk  at  Superfund  Sites:  A  National  Perspective
                                                Craig Zamuda, Ph.D.
                                                Environmental Scientist
                                   Office of Emergency and Remedial Response
                                      U.S.  Environmental Protection Agency
                                                   Washington, D.C.
ABSTRACT
  Reducing health risks to acceptable levels is a major objective
of the Superfund Program, with risk assessment a principle basis
for action. Procedures for conducting quantitative risk assessment
have been developed by the U.S. EPA and used to characterize
the risks at Superfund sites. This paper presents the results of a
survey of site-specific Superfund risk assessments, conducted by
the Office of Emergency and Remedial Response as part of the
RIIFS process for Superfund sites, to gain a national perspective
on the risk to human health  from inactive hazardous waste sites.
  This analysis attempts to capture the most significant risk from
Superfund sites. The basic analytical approach consisted of four
steps:
(1) estimate the mean individual cancer risk for hazardous sub-
stances frequently encountered at Superfund sites; (2) estimate the
number of people exposed per site; (3) estimate the total  number
of sites;  and (4) estimate the total risk to human health for inactive
hazardous waste sites in  the United  States.
  Results of the survey are presented in terms of six of the most
frequently occurring hazardous substances, the exposure pathways
posing the greatest risk to human health and the projected total
risk to human health from chemical exposure at inactive hazardous
waste sites, based on information obtained from the site-specific
assessments conducted at sites on the NPL. The results of the survey
and national  implications will be presented.

INTRODUCTION
  An objective  of  the  Superfund  program is to reduce to
"acceptable" levels the human health and environmental risks
caused by contaminants from inactive hazardous waste sites. Risk
assessment is the principal method for determining both the hazards
at those sites  before any  remedial activity and whether remedial
actions at those sites will reduce contaminant concentrations to
acceptable levels. Some form of risk assessment has been conducted
for most sites  where remedial  activities  have  been initiated.
Recently, the U.S. EPA conducted a survey of those risk assess-
ments to determine the level of risk that might be expected nation-
wide  from  uncontrolled hazardous  waste sites  due  to the
contaminants most frequently found at Superfund sites. A three-
part analysis was conducted. Initially, the risks associated with a
sample of Superfund sites were estimated. From these values, the
risks associated with all current and proposed NPL sites were esti-
mated. Finally, risks associated with the universe of all  inactive
hazardous waste sites were estimated. This survey was conducted
as part  of the U.S EPA study comparing overall human and
environmental risks reported in "Unfinished Business: A Compara-

56    POLICY ASSESSMENT
live Assessment of Environmental Problems"1, which compared
risks from a wide variety of environmental problems ranging from
ozone depletion to water pollution.
  This paper reports on the survey of risks attributable to current
and  proposed NPL sites,  as well  as  the universe  of inactive
hazardous waste sites,  by focusing on six of the most significant
carcinogens found at Superfund sites. Cancer risks are the focus
of this study because significant risk generally occurs at lower con-
centrations for carcinogens than for noncarcinogens and cancer
risk usually drives remedial activities at Superfund sites. Because
of the limited scope of this study (e.g., only six compounds out
of the hundreds found at Superfund sites), the estimated excess
cancer cases for the six  selected  chemicals may not accurately
predict the total cancer risk for Superfund sites.  In addition, neither
non-cancer nor environmental effects have been taken into account.
However, this study does give an indication of potential risks from
six  significant chemicals in one  significant  exposure  pathway,
groundwater.

PROCEDURES FOR THIS STl  DY
  A wide variety of exposures are  possible from hazardous waste
sites, including exposure to volatilized contaminants in air, inhala-
tion of wind- entrained dust, inadvertent soil ingestion, dermal
absorption from soil, dermal contact and ingestion of surface water
contaminated by leaching or running off from sites,  and con-
taminants leaching  into  groundwater.  Exposure  to surface or
groundwater can include dermal contact and inhalation of volatiles
as well as ingestion of drinking water. Out of the array of possible
exposure pathways, ingestion of contaminated groundwater was
associated with the overall highest  individual and population risks
for the sites included in  this analysis. Because consideration of
cumulative risks due to all exposure routes would be very difficult
and much less information is available  for the other exposure
routes, only risks due to groundwater ingestion were considered
in this study. This simplification eliminates consideration of some
exposure pathways that  contribute to  total risk  at inactive
hazardous waste sites; however, consideration of only groundwater
ingestion probably captures the majority of cancer risk at the sites.
  In order to describe the risks potentially due to inactive hazardous
waste sites, four steps were undertaken to estimate  individual and
population risks:

• Estimate mean individual cancer  risks per chemical based on
  available risk assessment documents and select the  most fre-
  quently occurring high-risk chemicals for further analysis
• Estimate the total number of sites to addressed  by this  study,

-------
  including sites listed on the National Priorities List and those
  not listed but still requiring attention
• Estimate the number of people exposed per site to hazardous
  levels of the most frequently occurring contaminants
• Multiply individual risk for each chemical times the exposed
  population times the number of sites to get overall cancer risk
  for the most frequently occurring chemicals of concern.

  This analysis attempts to capture the most significant risks from
Superfund sites. Because this analysis is limited to only six of the
literally hundreds of chemicals found at inactive hazardous waste
sites, total cancer risk from these sites is likely to be greater than
that estimated here. However, field experience at sites indicates
that the majority of cancer risks at most sites are due to a few
chemicals. In addition, only one of many possible exposure path-
ways has been analyzed; while the  most significant pathway may
have been considered, other potentially important pathways have
not been addressed. In fact, several sites used in this analysis did
not have any reported exposure to groundwater. On the other hand,
the estimates of the number of sites and exposed populations may
be overly conservative (i.e., too high). The actual number of cancer
cases in the population exposed to a carcinogen may also be too
large because  it is based on a 95th  percentile upper bound cancer
potency factor. The actual number of cases in the exposed popu-
lation is unlikely to be larger than those given but may well be
smaller. Nevertheless, this study provides a reasonable estimate,
given the quality and amount of data available, of the number of
excess cancer cases at NPL sites and all inactive hazardous waste
sites due to six significant carcinogens in the most frequently quan-
tified exposure medium.

Estimating Individual Cancer Risks
  Cancer risks reported in this study are based on information from
available  risk assessment documents,  usually  a  baseline public
health evaluation (PHE) or an endangerment assessment performed
as part of the RIF/FS process. These baseline risk assessments,
which  are based on procedures in the Superfund Public Health
Evaluation Manual2, generally assume that no remedial action
has taken place. The risk assessment documents identify the sig-
nificant exposure routes and chemicals of concern at a site and
often report excess lifetime cancer risks. Thirty-five risk assess-
ment or feasibility study documents were analyzed for this study.
Among the sites reviewed, the most frequently quantified exposure
pathway that was responsible for significant risk was groundwater
ingestion. Groundwater exposure risks were quantified at 20 sites.
Risks were much  less  frequently quantified for other exposure
routes.
  For each of the 35 sites, concentration and risk numbers were
abstracted for a maximum of five carcinogens for each exposure
pathway addressed in the risk assessment. Often, fewer than five
carcinogens were reported for a site.  For each  chemical, both
"plausible" maximum and best-estimate contaminant concentra-
tions and risks were averaged across sites for which they were
reported (i.e., sites with zero chemical concentration were not
included in the average).
  All compounds reported in more than two of the  assessments
are presented  in Table 1. "Frequency of appearance in PHEs"
refers to the number of sites at which the chemical was reported
according to the risk assessment documents (e.g., PHEs, RI/FSs)
for the 35 sites investigated. These values were used as the basis
for selecting the six chemicals that were the focus of this study.
"Percent frequency at sites  from CLP" refers  to the percentage
of sites where samples analyzed by the Superfund CLP indicate
the presence of those chemicals at the site. From the list of chemi-
cals compiled  from the site data, six chemicals responsible for the
highest cancer risks were selected for further analysis: arsenic, vinyl
chloride, tetrachloroethylene, trichloroethylene, 1,2-dichloroethene
and benzene. Generally, the chosen chemicals were associated with
both the highest concentrations and highest individual risks as well
as a high frequency of occurrence. The average risks for the six
selected chemicals were combined with the percent frequency that
these chemicals  were detected at sites (as  determined by CLP
analyses8 and the number of sites to extrapolate to the expected
cancer risks at sites nationwide.
                           Table 1
           Potential Carcinogens Found in Groundwater
     That Were Assessed for Risk in More Than One Superfund
            Public Health Evaluation/Feasibility Study



Trichloroethylene

Vinyl Chloride

Benzene

Tetrachloroethylen

Arsenic
1,2-Dichloroechane

Chloroform

PCBs

i, 1-Dichloroethene

. et y ene
1.1,2-Trichloroeth

N-nitrosodiphenyla
Frequency
PHEs a/

10

9

9

e 6


4

3

3

3


ane 2

nine 2
Fceq-
(from CLP)

27.9

6.8

30,4

22.6

'
9.4

27. B

a. i

12.9


4,6

6.9
Maximum
(UK/1)

40,000

5,500

1,622

12,000


9,000

2,800

27

72
6

1.500

90

Risk c/
-2
2 x 10
-1
2 x 10
-3
3 x 10
-2
2 x 10
-2
-2
2 x 10
-3
7 x 10
-3
3 x 10
-4
2 x 10
-2
-3
3 x 10
-7
9 x 10

(uR/O Risk e/
•2
20,000 1 x 10
-1
1,400 1 x 10
-4
200 3 x 10
-3
1,050 1 x 10
-3
-3
81 2 x 10
-4
98 2 x tO
-3
4.0 1 x 10
-5
5.6 2 x 10
-4
-4
23 3 x 10
-7
B x 10
  a/ Out of a total of 35 sites.

  b/ Site data Croa 20 NPL sites.  Averages are calculated with data from sites having chat
    chemical reported in public health evaluations only.
    chemical reported in public health evaluatioi
                                   .nly.
  d/ Site data froa 7 NPL. sites.  Averages are calculated
    chemical reported in public health evaluations only.
    chemical reported in public health evaluations only.
 Estimation of the Number of Sites
   After the average individual cancer risks were determined, the
 number of inactive hazardous waste sites at which people may be
 exposed had to be estimated. For NPL sites, the 951 current and
 proposed NPL sites (as of the July 22, 1987 NPL update) will
 represent NPL sites. The universe of inactive hazardous waste sites
 that may be associated with significant exposure includes both NPL
 and non-NPL sites. The number of sites that ultimately will be listed
 on the NPL has been estimated at 1800  in  a  study addressing
 CERCLA Section 301 (a)(l)(C)3. Non-NPL sites were estimated
 by subtracting the 1800 NPL sites from the approximately 25,000
 sites identified on the inventory list of potential and final CERCLA
 sites (CERCLIS) and assuming that two-thirds  of those sites are
 associated with no risk (as has been program experience), giving
 7,733 non-NPL sites and a total of 9,533 inactive hazardous waste
 sites.
   Other studies indicate that this number of sites may be too low.
 The Office of Technology Assessment has estimated the number
 of NPL sites at 10,000". Another U.S. EPA study estimated that
 Superfund ultimately will be used to clean up as many as 328,000
 sites5 ranging from NPL  sites to small-scale sites requiring only
 an immediate removal action to eliminate  risk at the site. If these
 estimates of the number of sites are more  accurate than the ones
 used in this study, then this  study may underestimate the actual
 risk from inactive hazardous waste sites However, whether the risks
                                                                                                  POLICY ASSESSMENT     57

-------
associated with each of the sites from the larger numbers of esti-
mated sites would be as high as the risks from the sites estimated
for this study cannot be  determined.

Estimation of Population Exposed per Site
  Estimates of the size of exposed populations are seldom availa-
ble from risk assessment documents because risk assessments tend
to focus on individual risk rather than population risk. Conse-
quently, site-specific population estimates from these documents
cannot  be used to estimate the average exposed population The
exposed population for this study was drawn from statistics in the
MRS data base, as cited in the CERCLA section 301(a)(l)(C)
study3
  Because this study is focusing on the groundwater exposure path-
way, the population exposed to groundwater contamination is the
important value. The mean number of groundwater users within
3 miles of NPL sites is 11,773. Ten percent of the preceding number
of people are assumed to be potentially affected by site contami-
nation because many of  the groundwater users within  the three
mile radius are likely to  be upgradient of the site,  too  far away
or not hydraulically connected to site contamination. Consequently,
the average population exposed to groundwater contamination at
NPL sites is estimated to be 1,177 NPL  sites are likely to be in
more populated areas than non-NPL inactive hazardous waste sites
because of the population factor included in the HRS scoring. Con-
sequently, the population exposed to groundwater contamination
at non-NPL sites is assumed to be half of that estimated for NPL
sites, or 589 people. Multiplying the total exposed populations at
the NPL and non-NPL sites by the number of each type of site
gives a  total of 6 7 million people.

UNCERTAINTIES AND CAVEATS
  Before the results are presented, the major uncertainties in this
study should be discussed. There are many assumptions built into
risk assessment numbers and inherent  in these extrapolation
methods. Taken as a whole, these assumptions place severe limi-
tations on the accuracy and applications of these projections. Major
assumptions  include:

• Site-specific risk assessment reports are both comprehensive and
  accurate.  In some cases these  were small-budget assessments
  based on very limited site data, and therefore the uncertainty
  in the individual risk numbers is high. In general, these numbers
  are probably conservative because of conservative assumptions
  in the exposure and toxicity assessments (e.g , use of future use
  scenarios, full lifetime exposure, upper-bound potency estimates'
  on which the risk analyses are based.
• Individual risk distributions and mean risks derived from the
  35 sample sites adequately represent the overall population of
  thousands of uncontrolled waste sites. This is a very small sample
  size, and it is highly unlikely to be representative given the great
  variability in uncontrolled waste site conditions and the sample
  site selection approach (the only criterion for inclusion was ready
  availability of a risk assessment report). This assumption also
  implies that non-NPL sites have similar risk levels to NPL sites
  because all sample sites examined were NPL sites.
• There will be no interventions to eliminate or reduce risk at these
  sites.  However, many sites are in the process of remedial action.
• The number of actual uncontrolled sites will not increase sig-
  nificantly over that projected  in  the  Section (301)(a)(l)(C)
  study3 (i.e., almost all existing sites already have been dis-
  covered). Other estimates indicate that the number of sites may
  be as  high as 328,000s.
• The mean exposed population  via groundwater contaminated
  by sites can be derived  from the estimated number of ground-
  water users within 3 miles. These numbers may not correlate very
  well.
• The entire exposed population either is equally susceptible (i.e.,
  no sensitivity variation) or the distribution of susceptibility is
  symmetric around the potency value used.
• There are no interactive effects (e.g., synergism) resulting from
  concurrent exposures to multiple chemicals in the waste or from
  sources other than uncontrolled waste sites.
• The distributions of individual risk and exposed population are
  independent (i.e., there is no association between population size
  and risk level). If these distributions are skewed (likely) and non-
  independent  (possible), the population risk estimates could be
  significantly  in error.

  Clearly, the number and nature of assumptions required to esti-
mate population risks limits the confidence in the estimated results;
the more assumptions that have to be made the less confidence
can be placed in the results.  For example, risk numbers for the
35 sites require no extrapolation and  are more reliable than the
estimated cancer cases for the universe of inactive hazardous waste
sites, where the number of sites is based  on a large number of
assumptions.

STUDY RESULTS
  The estimated annual numbers of excess cancer cases due to the
six selected carcinogens from the groundwater pathway at the 35
sites are reported in Table 2.  These estimates assume an exposed
population of 1,177 at each site as determined by the methodology
above. These estimates refer  only to the 35 sites investigated for
the analysis, not to the "typical" Superfund site or the universe
of inactive hazardous waste sites.  Vinyl chloride was associated
with the highest number of excess cancer cases, with 30 and 20
excess cases for  maximum and best-estimate situations, respectively.
                           Table 2
            Projected Annual Exeea Cancer Cam for
             Selected Chemicals al 35 In Sites Having
                   Public Health Evaluations'
                            Cancer C*uซ*
                                         B*st-EstlซMtt> Excels
                                           C*ncซr Cซsซi
Trichloroethylene
Vinyl Chloride
Benzene
Tetraehloroathylene
Arsenic
1 , 2-Dlchloroelhane
}
30
0.5
1
1
1
2
20
0.
0
0
0


.OS
.1
.J
.1
  If Assuming an average of  1,177 people exposed Co ground water at each
    site.  See teat (or ewchod.  Annual nuvber of cases represents the
    lifetime nuabar of projected cases divided by an average 70-year
    lifetime. Values are rounded to one significant figure, which is the
    level of precision available fro* the risk estimates.
  The estimated annual numbers of excess cancer cases due to the
six  selected carcinogens  from the groundwater pathway at all
current  and proposed NPL sites are reported in Table 3. These
values represent the estimated number of excess cancer cases among
the estimated 1,177 people exposed at each of the 951 current and
proposed NPL sites as determined by the methodology above. The
maximum and best-estimate excess cancer cases per year for all
NPL  sites were 200 and  100 excess cases, respectively, for vinyl
chloride. The relatively  high  mean concentrations and cancer
potency factor for vinyl chloride account for its high excess cancer
cases.
  Arsenic has the second highest maximum and the third highest
best-estimate excess cancer cases. Arsenic is found in very low con-
centrations relative to the other five chemicals chosen for this study;
however, it is found at a very high percentage of the sites and has
a very high potency factor. The remaining chemicals are found
58    POLICY ASSESSMENT

-------
in relatively high concentrations or at a relatively large number
of sites, which accounts for their significance.
                            Table 3
            Projected Annual Excess Cancer Cases for
      Selected Chemicals at Current and Proposed NPL Sites1
                          Maximum Excess  Best-Estimate Excess
                           Cancer Cases       Cancer Cases
Trichloroethylene
Vinyl Chloride
Benzene
Tetrachloroethylene
Arsenic
1 , 2 -Diehloroethane
90
200
10
70
100
30
40
100
1
4
30
3
 I/ Assuming an average of 1,177 people exposed Co ground yater at each of
    the 951 sites.  See text for method.  Annual number of cases represents
    the lifetime number of projected cases divided by an average 70-year
    lifetime.  Values are rounded to one significant figure, which is the
    level  of precision available from the risk estimates.
   Table 4 presents annualized excess cancer cases for the top six
 chemicals extrapolated to the entire potential universe of inactive
 hazardous waste sites. These values represent the estimated number
 of excess cancer cases among the estimated exposed population
 of 6.7 million as determined by the methodology above. Maximum
 excess cancer cases per year for all sites ranges from 1000 for vinyl
 chloride to 90  for benzene. Best-estimate values range from 700
 for vinyl chloride to 9 for benzene.
                            Table 4
             Projected Annual Excess Cancer Cases for
              Selected Chemicals at Entire Universe of
                     Hazardous Waste Sites1
                           Maximum Excess   Best-Estimate Excess
                           Cancer Cases       Cancer Cases
Trichloroethylene
Vinyl Chloride
Benzene
Tetrachloroethylene
Arsenic
1 , 2 -Diehloroethane
if Total population potentially
500
1000
90
400
800
200
exposed to
300
700
9
20
200
20
inactive hazardous waste sites
    is estimated to be 6,7 million.  See text for method.  Annual  number of
    cases represents the  lifetime number of projected cases divided by an
    average 70-year lifetime.  Values are rounded to one significant
    figure, which is the  level of precision available from the risk
    estimates.

 CONCLUSIONS

   The purpose of the study reported here was to estimate cancer
 risks from  inactive hazardous waste sites. The results should be
 balanced against the inherent  uncertainties  associated with the
 limited number of non-randomly selected sites, the limited exposure
 pathways investigated, the limited number of chemicals analyzed,
 and the lack of noncancer data addressed by this study. As noted
 above, however, the cancer risks tend to occur at lower concen-
 trations than noncancer risks and tend to be caused by a relatively
small number of the same chemicals at each site. Consequently,
this study, while very limited, does provide important risk infor-
mation about sites.
  Much of the cancer risk at Superfund sites is due to only a few
chemicals. Although not quantified, much of the carcinogenic risk
at the sites investigated appears to result from only a few chemi-
cals at each site. This conclusion supports the Superfund policy
of selecting indicator chemicals for sites having large numbers of
chemicals and indicates that if the proper subset of the chemicals
found at sites is addressed in the risk assessment, the assessment
will include all of the significant risks at the site.
  Annual excess cancer cases for six significant carcinogens in the
most frequently quantified exposure pathway are estimated at 500
(maximum) and 200 (best-estimate) for current and proposed NPL
sites. For the universe of all inactive hazardous waste sites the excess
cancer cases are estimated to be approximately 3000 (maximum)
and 1200 (best-estimate).
  Recently, the U.S.  EPA conducted an agency-wide survey of
environmental problems that was summarized in the ' 'Unfinished
Business" report1. Thirty-one areas were investigated  ranging
from indoor radon exposure to pesticide residues on food. Among
the problem areas included were indoor air pollution, carbon
dioxide and global warming, biotechnology, mining waste, inactive
hazardous waste sites, active hazardous waste facilities, storage tank
releases and point source water pollution. Cancer and noncancer
risks, as well as environmental and welfare effects were estimated.
Cancer risks for some problems  such as indoor radon exposure
(5,000 to 20,000 excess cancer cases) and pesticide residues on food
(6,000 excess cancer cases) were estimated to be higher than cancer
risks estimated from inactive hazardous waste sites. Other problem
areas such as drinking water (400-1000 excess cancer cases) active
hazardous waste sites (fewer than 100 excess cancer cases), and
releases from  storage tanks (less than one excess cancer case per
year) were estimated  to  have lower cancer  risk  than  inactive
hazardous waste sites. Like the risks from inactive hazardous waste
sites,  all of these estimated risks are subject to considerable un-
certainty. However, the results of this project are intended to serve
only as a guide to broad, long-term priority setting. The U.S. EPA's
response to a risk might involve one or more of the following ac-
tivities:

• Conducting research to understand the problem and/or develop
  methods of control
• Disseminating information and educating the public
• Initiating or increasing program activity, such as issuing regu-
  lations, writing permits or enforcing regulations and  permits
• Asking or helping others (e g., Congress, state and local govern-
  ments, individual citizens) to legislate or take appropriate action
  or even
• Taking no action, where the risk is low

  Significant  improvement in the accuracy of the cancer risk
estimates from Superfund  sites will be possible as additional
information is made available from Superfund remedial investi-
gation/feasibility documents. In general, developing new data on
carcinogenic substances and human exposure to carcinogens in the
environment takes a considerable amount of time. However, the
Superfund program is actively accumulating such information not
only for carcinogens but also for non-carcinogens. A better under-
standing of the magnitude and sources of risk to human  health
resulting from uncontrolled hazardous waste sites will allow the
Superfund program to focus its resources on those compounds and
exposure pathways that account for the most significant risks at
sites.

REFERENCES
1. U.S. EPA. "Unfinished Business: A Comparative Assessment of Envi-
   ronmental Problems-Overview Report." Office of Policy, Planning, and
   Evaluation,  Washington, D.C. Feb. 1987.
                                                                                                    POLICY ASSESSMENT     59

-------
 2. U.S. EPA. Superfund Public Health Evaluation Manual. Office of        4. Office of Technology Assessment. "Superfund Strategy." Washington,
   Emergency and Remedial Response, Washington, D C. Oct. 1986. EPA           D.C.,  OTA-ITE-252, April 1985
   540/1-86-060.
 3. U.S. EPA. "Extent of the Hazardous Release Problem and Future Fund-        5.  Hayes, D.J. and MacKerron C.B. "Superfund II: A New Mandate."
   ing Needs CERCLA Section 301(a)(l)(C) Study." Office of Solid Waste           Washington, O.C., A Bureau of National Affairs Special Report. Envi-
   and Emergency Response, Washington D.C. Dec. 1984.                      ronment Reporter. 77(42), 1987, p.13.
60     POLICY ASSESSMENT

-------
                                 Risk  Assessment in Superfund:
                                        Policies  and  Procedures

                                                     Sherry Sterling
                                                Craig Zamuda, Ph.D.
                                      U.S. Environmental  Protection Agency
                                 Office of Solid  Waste and Emergency Response
                                                    Washington, DC
ABSTRACT
  As a result of the reauthorization of Superfund, there have been
substantial changes in the manner in which the EPA addresses risk
assessment  in  the Superfund remediation process. This paper
reviews the policies and procedures for conducting risk assessments
at Superfund sites. Specifically, this paper reviews the manner in
which risk assessment will interface with the other aspects of the
Superfund process by focusing on policy documents such as the
reviewed NCP, the RI/FS and other procedures and guidance in
the area of risk assessment.
  Manuals  such as the Superfund Public Health Evaluation
Manual and the Superfund Exposure  Assessment Manual are
important in this risk assessment process.  Other resources that have
been developed to analyze public health risk at Superfund sites and
to develop health-based cleanup levels for remedial action, such
as the Risk Assessment Information Directory  and the Public
Health Risk Evaluation Database, Health Effects Assessment docu-
ments and lexicological profiles, will be introduced with regard
to their utility in the risk assessment process. In addition, this paper
will provide information concerning human resources available at
U.S. EPA Headquarters  in the Offices of Emergency and Remedial
Response and Waste Programs Enforcement.


INTRODUCTION
  The protection of human health and  the environment are key
requirements in selecting the remedy for sites under CERCLA.
SARA addresses these requirements in  several ways. For exam-
ple, Section 121 of SARA specifically requires the selection  of a
remedy that is protective of human health and the environment.
In addition, Section 104 has been substantially amended to include
certain health-related authorities which the U.S. EPA shares with
the Agency for Toxic Substances and Disease Registry (ATSDR).
These responsibilities include conducting health assessments at all
sites proposed on the NPL, conducting epidemiologic studies and
establishing disease and exposure registries.
  The Superfund risk assessment process is part of a larger process
developed by the Agency to arrive at decisions related to the selec-
tion of a remedy at a Superfund site. This larger process, namely
the RI/FS, includes a number of phases: scoping, site characteri-
zation, development of  alternatives, initial screening of alterna-
tives, treatability studies and detailed analysis of the alternatives.
Work on the risk assessment is begun in the scoping phase of this
process; however, the bulk of the risk assessment is conducted in
the site  characterization phase and in the alternatives screening
process.
RISK ASSESSMENT IN SUPERFUND
  The basic steps in the Superfund risk assessment process are
based upon those steps outlined  in the National  Academy of
Sciences  study Risk Assessment  in the Federal Government:
Managing the Process1 that has been adopted by the U.S. EPA.
These steps are: hazard identification, dose-response evaluation,
exposure assessment and risk characterization.
  Superfund risk assessment in the RI/FS context is best described
as an iterative process. First, a baseline risk assessment is conducted
to determine the risks to human health and the environment if no
remedial action is taken at the site. Subsequent iterations of the
risk assessment methodology, as outlined in the Superfund Public
Health Evaluation Manual2-,  assess the  risks from  various
remedial alternatives under consideration compared to the base-
line risk assessment.
  A baseline risk assessment is an analysis  of site conditions in
the absence of remedial action. The assessment provides an under-
standing of the nature of chemical releases from the site, the path-
ways of human exposure, the degree to which such releases violate
applicable or relevant and appropriate requirements and a measure
of the threat to public health as a result of releases. The informa-
tion developed in the baseline assessment provides input to develop
and evaluate remedial  alternatives. In  addition,  the  baseline
assessment  satisfies the NCP requirement to complete a detailed
analysis  of  the no-action alternative, including an evaluation of
public health impacts3.
  Developing performance goals  for remedial alternatives is
another important phase of risk assessment at Superfund sites. This
phase of the risk assessment process builds on information  collected
and evaluated in the baseline risk assessment phase.  In this phase
the  risks from the various alternatives are estimated in  order to
determine the level of protectiveness that would be afforded by
the  various remediation alternatives.
  SARA requires that the selected remedy at least attain levels of
"applicable or relevant and appropriate" requirements (ARARs).
Basically, ARARs can be broken into three  categories: chemical-
specific, action-specific and location-specific. The most relevant
of these types of ARARs for purposes of developing acceptable
levels of exposure are the chemical-specific ARARs. Potential
chemical-specific ARARs may include Maximum Contaminant
Levels (MCLs), Maximum Contaminant Level Goals (MCLGs) and
certain State standards. The NCP describes how ARARs are to
be used in determining acceptable levels for  cleanup. In addition,
the Agency has issued the "Interim Guidance on Compliance with
Applicable  or Relevant and Appropriate Requirements"4.
  There may be cases where ARARs do not exist for all chemicals
                                                                                             POLICY ASSESSMENT    61

-------
at the site. In these circumstances, the results from the risk assess-
ment are used together with whatever ARARs are available to set
cleanup levels.
  Because of circumstances that are found at the site, ARARs alone
may not be protective of human health and the environment. For
example, when the total carcinogenic risk from all chemicals at
the site is calculated using the ARARs, the estimated risk may not
be considered protective due to the interactions of all of the chem-
icals present. For this reason, it is important that the total risk from
all chemicals at a site be considered before the final cleanup levels
are defined.

GUIDANCE DOCUMENTS
  Details concerning these general components of the U.S. EPA
risk assessment process have been described in the Superfund Public
Health Evaluation Manual (SPHEM). This manual, which was
published in October 1986, provides a basic, but relatively compre-
hensive framework for the development  of risk  assessments for
Superfund sites.
  The SPHEM, which was developed for Superfund-led sites, may
also be used in enforcement cases.  Although the level of effort may
differ in an enforcement-lead situation,  the basic process  is the
same. For admininstrative and judicial enforcement actions  under
Section 106 of CERCLA, an endangerment assessment is neces-
sary to justify the enforcement action. The risk assessment process
is used to make the determination of endangerment to public health
or  the environment.
  To supplement the basic risk assessment manual, the Office of
the U.S. EPA Emergency and Remedial  Response (OERR) is in
the process of finalizing an Exposure Assessment manual which
will provide detailed information concerning exposure evaluation.
In addition, OERR and the U.S. EPA Office of Waste Programs
Enforcement (OWPE) are working on guidance for  conducting
environmental  risk assessments. These  guidances  have been
developed to answer many of the questions which are commonly
raised by the U.S. EPA regions, responsible parties and contrac-
tors during the course of the RI/FS regarding risks to humans and
the environment resulting from exposure at Superfund sites.
  In addition  to these guidance documents, there are a  number
of  resource documents  prepared by OERR that support risk
assessment activities. These include the Superfund Risk  Assessment
Information  Directory5  that provides basic information and
descriptions of databases,  models,  publications and  human
resources that should be useful in conducting risk assessments.
  Another source of information is the Agency's Integrated Risk
Information System  (IRIS) database which  contains results  of
carcinogenic bioassays, dose-related responses, toxicity levels, refer-
ence doses and other parameters necessary to conduct risk assess-
ments  at  Superfund  sites. Designed as  a electronic loose-leaf
notebook, IRIS is accessed through commercial E-mail lines. The
system is organized on a chemical basis and the user  can call  up
a chemical by name and review all material pertinent to it.
  In addition to IRIS, the Superfund program has developed the
Public Health Risk Evaluation Database (PHRED), a personal
computer software package designed to provide chemical, physi-
cal and toxicological data, health-based standards and criteria for
over 400 chemicals that may be found at Superfund sites. Together,
PHRED and  IRIS constitute important sources for up-to-date
Superfund risk assessment information.
  The U.S. EPA Health Effects  Assessments (HEAs) provide a
valuable source of information on toxic chemicals typically found
at Superfund sites. The HEAs summarize and evaluate informa-
tion relevant to a preliminary assessment of adverse health effects
exposure levels whenever sufficient data are available. In total, over
100 individual HEAs are available for specific chemcials or chem-
ical groups.
  The SARA amendments also require ATSDR to develop Toxi-
cological Profies on at least 25 chemicals each year,  starting with
the most toxic chemicals that are commonly found at Superfund
sites. The first 25 profiles had to be completed by ATSDR 1 year
after the enactment of SARA (Oct.  17, 1986). These profiles con-
tain a summary and interpretation of available  toxicological
information and epidemiological evaluations on a  substance, a
determination of whether adequate information exists on the health
effects of the chemical and an identification of toxicological testing
needs.

HUMAN RESOURCES
  OERR and OWPE both maintain staffs to provide technical sup-
port for risk assessment  under Superfund. The objectives of the
Toxics Integration Branch, OERR  include:

• Developing guidance for analyzing health and environmental risk
  at Superfund sites
• Providing technical information and  quality assurance review
  of Superfund risk assessment activities in the U.S.  EPA regions
• Providing a focal point within OERR for information on risk
• Providing risk assessment training to U.S. EPA  regions.

  Similarly, the staff of the Health  Sciences Section in  the
CERCLA Enforcement Division of OWPE provides information
and answers questions concerning risk assessment for both human
health and  the environment and on  national policy concerning
Superfund risk assessment.  Information or questions concerning
specific Superfund sites should be directed to the appropriate U.S.
EPA  regional  office for that site.

CONCLUSION
  The guidance development efforts of the U.S. EPA  assist in
formulating the science policy basis of risk assessment; however,
additional research efforts are needed to address the  uncertainties
in quantitative assessments. As a result of requirements in SARA,
Superfund resources can  be used to conduct hazardous substances
research. This has enhanced U.S. EPA  research efforts  and will
broaden  our  understanding, not only of the physical-chemical
properties and behavior  of hazardous substances in the  environ-
ment, but also of how hazardous substances cause adverse health
effects. In addition to U.S. EPA research efforts, other Superfund
research efforts include those of the ATSDR of the Department
of Health and Human  Services (e.g.,  epidemiology studies at
Superfund  sites) and the National Toxicology Program (e.g.,
toxicity testing of Superfund hazardous substances). These efforts
are all part of a growing effort to improve the scientific basis of
risk assessment.

DISCLAIMER
  This paper has not been subjected to U.S. EPA review and there-
fore does not necessarily  reflect the views of the U.S.  EPA Agency
and no  official endorsement should be inferred.

REFERENCES

1. National  Academy of  Sciences, Risk Assessment in the Federal
   Government: Managing the Process, National Academy Press, Washing-
   ton. DC,  1983.
2. U.S.  EPA,  Superfund  Public  Health  Evaluation Manual,  EPA
   540/1-86/060, U.S. EPA. Washington. DC,  1986.
3. Federal Register, 50, No. 224, Nov. 20,  1985. 47912-47979.
4. Porter, J. W., "Interim Guidance on Compliance with Applicable or
   Relevant and Appropriate Requirements," Office of Solid Waste and
   Emergency Response Directive 9234.0-05, July 9, 1987.
5. U.S. EPA, "Superfund Risk Assessment Information Directory." EPA
   540/1-86/61, U.S. EPA, Washington, DC, 1986.
 62    POLICY ASSESSMENT

-------
                                 Selecting  A Contract  Laboratory

                                                      Craig W. Rice
                                            Hittman Ebasco Associates Inc.
                                                   Columbia, Maryland
ABSTRACT
  Since 1980, the environmental laboratory business has enjoyed
a period of rapid expansion. Passage of CERCLA, SARA, recent
amendments to the Safe Drinking Water Act, proposed revisions
of the National Primary Drinking Water Regulations and closer
scrutiny by the media of the ways in which hazardous substances
are stored, used and disposed of, have heightened public awareness
of chemical pollution problems and stimulated further efforts to
regulate these activities. This process has generated a large demand
for contract laboratory services and  led to the formation of many
new companies.
  For those inexperienced in laboratory evaluation, choosing among
the available laboratories can be a formidable challenge. This paper
is presented in an attempt to educate managers, supervisors and
private individuals as to how best to proceed in locating, evaluating
and ultimately selecting a contract  laboratory to provide analytical
services best suited to specific  project requirements.

INTRODUCTION
  In July 1987, state governments under authority of the U.S. EPA,
began to implement the first phase of the 1987 Amendments to the
Safe Drinking Water Act of 1974. Although this regulation and pro-
posed amendments to the National Primary Drinking Water Regula-
tion to be phased in over the next  5  years directly affects only the
potable water supply industry, it is evident that expanding govern-
mental regulation has and will continue to affect every private and
commercial enterprise that deals with  the nation's fresh water supply
as well as chemical wastes and their treatment and disposal. A
number of laws have been passed in recent years which regulate the
use and  disposal of potentially hazardous  chemicals, delineate
requirements for testing the work environment, provide workers with
right-to-know legislation and promulgate strict permit requirements
for industrial discharge. Such legislative fervor seems to reflect a
desire on the part of the general public to become more aware of
the unseen hazards of chemical pollution. Many of these laws require
extensive chemical analysis.
  In response to the increasing demand  for analytical laboratory
services, many new companies have been formed. Other companies
have developed their own analytical capabilities for laboratory testing
of waste streams and effluents,  developing and evaluating pretreat-
ment programs, or determining the proper classification of poten-
tially hazardous materials and providing technical advice for proper
disposal. In view of the diverse nature of laboratory services, their
potential application to a wide range of problems  and the sheer
number of companies involved in providing analytical  services in
a rapidly expanding market, selecting the "best" laboratory for a
given purpose has become more difficult and confusing.
  The purpose of this paper is to present a detailed plan to deter-
mine how to choose an independent contract laboratory. The paper
begins with a survey of the various reasons one might wish to use
such services and proceeds to consider expense, qualifications, data
validation and impartiality. A list of comparison criteria will be
discussed. Although every criterion may not apply in each specific
case, following the suggested plan will allow a manager, supervisor
or private individual to develop a set of criteria with which to evaluate
potential contract laboratories in a way that is meaningful to the
specific project requirements.

WHY DO IT YOURSELF?
  For private individuals, laboratory testing at home is not feasi-
ble. Lacking considerable wealth and extensive training, today's base-
ment scientist is ill-equipped for environmental monitoring. Many
of the common drinking water contaminants such as lead or
trichloroethylene are considered health risks at concentrations below
30 iig/l, a level far lower than the detection limit of any simple test
kit.  Specialized  analytical instruments and meticulous  sample
preparation procedures must be employed in order to detect and
accurately quantify such contaminants. Testing for pesticides or the
presence of radon gas, both of which have been identified as carcino-
gens with the potential  to cause serious household contamination
problems, requires specialized sampling techniques, expensive testing
equipment and experience in data interpretation.
  Even large manufacturing companies often cannot afford to ade-
quately equip an in-house environmental testing laboratory. In ad-
dition, the time  spent in  hiring  experienced chemists  or training
employees to perform these tests may not be cost effective unless
protracted use of these services is anticipated. Because they perform
hundreds of analyses on a daily basis, contract laboratories should
be able to provide specialized testing services to include expertise
in sample collection and  preservation, experience in chemical analysis
and competent data interpretation at relatively low per-sample cost.
  Companies with in-house testing laboratories may need to use an
independent  laboratory  because  of  equipment  maintenance
problems, to relieve a temporary over-capacity situation or as a check
on the quality of data produced in-house. In addition, use of an
independent laboratory lends a  dimension of  impartiality to the
results.  Just as the books of commercial lending institutions must
be audited periodically, independent surveys of public opinion are
conducted  for  marketing purposes and  advertising  claims are
scrutinized by outside agencies, so the perception of independent
verification as inherently more objective and free from conflicts of
interest can be an essential element in support of analytical data.

                     SAMPLING AND MONITORING    63

-------
"Tested by an independent laboratory" is a common advertising
claim whose purpose is to present results and conclusions as more
credible while limiting  the  potential  liability  of the client.  In-
dependence is rarely complete, however, since  the client pays for
these services and must rely on  the technical competence of the
laboratory.

GETTING STARTED
   The first step in selecting a contract laboratory is to compile a
list of candidates. The local  telephone book is often the best place
to start. A major metropolitan telephone yellow pages section may
contain as many as 100 or more  individual listings for various
laboratory services. Narrow your focus by looking under sub-
headings such as analytical,  bacteriological, research and develop-
ment or materials testing as appropriate. Note  areas of  specializa-
tion. The advantage of obtaining  numerous references from a single
source is somewhat offset by the lack of detailed information by
which the list  may be narrowed. Check the advertisements in
technical  publications,  especially those that directly relate to or
specifically target your  industry  or corporate specialty.
   Many government agencies will respond to requests for names and
addresses of laboratories  certified by them.  Examples include
U.S. EPA,  USATHAMA, U.S.  Army Corps  of Engineers, state
environmental agencies, state and local water resource agencies and
health departments.
   One of the most valuable sources of information may be recom-
mendations from associates whose specific needs may be similar to
your own. This list may be short, but all favorable recommendations
will warrant further consideration. From sources outlined above,
eliminate those candidates that do not have the appropriate analytical
capabilities  and then begin  to compare the remaining companies
based on as many of the evaluation criteria described in the following
section as practicable. In general, the final selection should be based
on the quality and timeliness of the work and, of course, on price.

COMPARISON CRITERIA
   By following the basic approach outlined in this section, enough
information will be available to make a meaningful comparison of
various candidate laboratories. Much of the information should not
be proprietary. Laboratories that  hesitate or refuse to provide clear
answers should be eliminated from further consideration. However,
it  will not be possible to adequately evaluate all of the comparison
criteria by verbal interaction alone, so plan to visit the facilities of
as many of the final candidates  as time will allow.

Analytical Methods
   Sample preparation and analytical procedures used by a contract
laboratory should be based upon established methods published in
standard  reference materials  and accepted by  the appropriate
regulatory agency. References  such  as  those compiled  by the
U.S. EPA, APHA, ASTM, SEMI, etc., are well-known throughout
the industry. The degree to which the implemented procedures cor-
respond to a  published method can be critical if the project involves
regulatory oversight, permit application or eventual litigation. A con-
tract laboratory  should be  able to provide a detailed Standard
Operating Procedure based  on a published method  for each test.
Obtaining this information is the only way to determine the degree
to which data from different sources are comparable.

Quality Assurance
   A vast amount of material has been written on this subject, and
a  detailed discussion  is outside the  scope  of this  paper. All
laboratories  should have a detailed written QA/QC program. As
a minimum, the program should address chain of custody and sample
control, routine calibration and  maintenance of equipment.  The
laboratory should be able to supply precision and accuracy data from
the analysis  of duplicate samples,  spikes and  standard reference
materials.
Staff Qualifications
  Candidate laboratories should be staffed by scientists and tech-
nicians whose education is appropriate to the tasks they will per-
form. A Laboratory managed by a history or Spanish major may
not be able to provide the expertise of one managed by a Ph.D. in
organic chemistry, but other factors may be important.  Examine
not only the quality and quantity of higher education, but also the
level of experience in  performing the analysis. If a complex matrix
such  as  industrial sludge  or  soils with  high organic content is
anticipated, ensure that the staff has experience in working with these
materials. Resumes should be available upon request.

Sample Handling
  Contract laboratories should provide complete sets of containers
with appropriate preservatives, chain of custody forms and instruc-
tions for collection  and  delivery of samples to  their faculty.
Sometimes this service is included in the price of the analytical work,
often it is billed as a  separate item. Bottles should be pre-cleaned
and supplied with labels. A comparison of laboratories should focus
on availability of complete sampling kits and ease of return delivery.

Data Turnaround
  The elapsed time between sample receipt by the contract laboratory
and delivery of the final report should be firmly established. Most
laboratories offer a "normal" sample turnaround time with addi-
tional premiums for rush or emergency service.  If time is a critical
factor, explore  the  possibility of obtaining preliminary results
verbally with a formal report to follow. The best way to evaluate
sample turnaround is  to compare what the laboratory advertises to
what their clients actually experience. Ask for a list of clients in order
to verify data turnaround time.
Communication
  The impression of scientists as infallible hierarchs who speak an
arcane dialect and engage in mysterious rites to which the uninitiated
are denied admission is, for the most part, a fallacy. Staff members
should be able to communicate ideas and information in a way that
is easily understood by non-scientists. If not, their advice and recom-
mendations may be confusing or unintentionally misleading.

Confidentiality
  Contract laboratory services should be  confidential. A client's
name should not be given without consent. All data should be con-
sidered proprietary, and they should not object to signing a state-
ment of confidentiality. Find out  who has access to the data and
be sure to ask if all work will be  done in-house. Subcontracting
analytical work is a common practice for a number of acceptable
reasons such as equipment  maintenance problems, an over
-------
laboratory performance information. These standards consist of a
series of sealed ampoules containing one or more compounds for
quantitation. Select  a set that best  represents  the project's test
requirements.
  Performance evaluation standards can be ordered directly from
the U.S. EPA at little or no cost. When they arrive, remove the labels
and number the vials for identification.  Keep accurate records on
the number assigned and the corresponding label.  Save the tabulated
true values provided with the kit and send the numbered ampoules
and the preparation instructions to each of the final candidate
laboratories.
  Based on the results of performance standards, a further evalua-
tion can be made of analytical accuracy as well as turnaround time
and the quality of the final report. After a final review of these results
and as many  of the  above comparison criteria  as  possible,  an
appropriate choice should  be obvious.
                                                                                        SAMPLING AND MONITORING     65

-------
                      The  Use  of Short-Term  Bioassays to Assess
                      Clean-Up  Operations  of  Sites Contaminated
                                       With  Hazardous  Wastes

                                                     Phebe Davol
                                                 A. T.  Kearney, Inc.
                                                 Alexandria, Virginia
                                                  Kirby  C.  Donnelly
                                               Texas A&M University
                                                College  Station, Texas
                                                    Kirk W. Brown
                                               Texas A&M University
                                                College  Station, Texas
ABSTRACT
  The fate of organic constituents from soils contaminated with
hazardous industrial waste was monitored for a 4-year period using
a microbial mutagenicity assay and GC/MS/DS analysis. Soil,
plants and runoff water were periodically collected and solvent was
extracted and examined for bacterial mutagenicity. While chemi-
cal analysis was useful in the identification of hazardous consti-
tuents in the environmental samples, the bioassay measured  the
genotoxic potential of components in these complex mixtures.
Information from this study suggests that a bioassay-directed chem-
ical analysis may be an effective tool in determining the potential
lexicological hazard from exposure to hazardous organic con-
stituents and the effectiveness of remedial actions to permanently
and significantly reduce the mobility and toxicity of contaminants
from  a site contaminated by hazardous wastes.

INTRODUCTION
  Remedial actions selected for a Superfund hazardous waste  site
must be sufficient  to protect human health and the environment.
For the proposed action, one must sufficiently document that there
will be minimal long-term effects to the environment from any
residual contamination; one must also show the proposed cleanup
will protect human health, both  during  and after the remedial
activity. To what standards the sites must be cleaned has not been
determined. One alternative to relating monitoring at a hazardous
waste site to risk assessment is to compare levels of contaminants
detected  to currently  existing air or  water quality  standards.
[(Maximum Contaminant Levels (MCLs), Maximum Contaminant
Level Goals (MCLGs) or Alternate Concentration Limits (ACLs)
have been suggested for groundwater). Unfortunately, standards
and MCLs exist for only a few chemicals; and once the standard
has been exceeded, remedial action may be ineffective  or cost-
prohibitive. Preferentially, the goal is to permanently and signifi-
cantly reduce the mobility, toxicity and volume of waste before
groundwater is affected. Cleanup standards of SARA Section  121
include an emphasis on risk reduction  through detoxification or
destruction of wastes. In order to determine whether a waste  has
been detoxified, a biological analysis of environmental samples  col-
lected from a contaminated site should be conducted to provide
a more accurate assessment of the effectiveness of the remedial
activity.
  Human exposure to environmental mixtures may occur as  chem-
icals are volatilized or released into the atmosphere, absorbed by
particulates, leached into groundwater, runoff into surface water
or translocated into plants. Once in the environment, complex mix-
tures  can be transformed or degraded  by chemical or biological
action in the air, soil or water. The lexicological properties of com-
ponents of a complex  mixture may be significantly altered by
environmental transformations, and by their synergistic, antagonis-
tic or additive interactions.
  Thus, for a procedure to be a useful tool for monitoring poten-
tial carcinogens in  environmental samples,  it must be simple,
economical and have the ability to detect metabolites of environ-
mental transformations  and the various toxic interactions. Perhaps
the best compromise between simplicity and accuracy is provided
by the Salmonella microsome assay developed by Ames and co-
workers'.  The  objective of this  research was to  monitor soil,
plant, and runoff water samples from hazardous industrial waste
contaminated soils for the presence of mutagens and potential car-
cinogens. The procedure measures the ability of a sample to induce
reverse mutations in a specially constructed strain of Salmonella
typhimurium. Generally, the test is considered to be 85 to 95%
efficient  in  detecting  mammalian  carcinogens   as  bacterial
mutagens2; although when a broad range of chemical classes are
included, the accuracy may be as low as 58%'. However, if based
on significant positive carcinogens (chemicals which induce tumor
formation in both sexes of two species), the accuracy of the test
may be as high as 95ro-"
  These data indicate that for potential carcinogens with the most
direct implication for human health effects, microbial  bioassays
are an effective monitoring tool. Thus, microbial  bioassays can
be used to evaluate  the  mutagenic activity of environmental sam-
ples as an indication of their potential to induce mutagenic damage
in the human population.  In humans, mutagenic compounds af-
fecting somatic cells may induce cell death, cancer, aging and heart
disease" while mutations in germ cells may result in birth defects,
sterility and abortions7 (Figure 1).
                 Gซrm Ctlll
                 (Reproductive)
       Mullllon
    (AJiemkxi In DMA or
    CylopUlmlc
                              Domininl (—\
                              Mutilloni '—•/
                              Recrnive
                              MuUdoni
Somปtlc Celts
(Non-Rrproduc(lvt)
                               o
• Binh Defects
• Geneoc Disease
• Abortions
• Sterility
Expressed is Genetic
Disease in Future
Oenerancm
 CeUDeuh
 Cmer
 Aflat
 Heart Discue
 Other fflnco
                        Figure 1
   Potential Effects of Environmental Mutagens on Human Cells
66    SAMPLING AND MONITORING

-------
MATERIALS AND METHODS
  Two petroleum-based sludges, wood-preserving bottom sediment
(PENT S) and combined API separator/slop oil emulsion solids
(COMBO) refinery waste were collected in 55-gal barrels. (A more
detailed chemical description of these wastes is given by Brown
et al.[8]. Two  soil  series, Weswood silt loam (Fluventic
Ustochrept) and Bastrop clay loam (Udic Paleustalf), representing
diverse soil textures were utilized.
  The waste-treated soils were packed in wooden boxes 60 cm  x
45 cm x  17 cm and subjected to one hour of simulated rainfall
at an intensity of 8.9 cm/hr immediately after, 180, 360, 540 and
1000 days after waste application. Runoff water was collected in
amber glass bottles,  passed  through an  XAD resin column and
extracted with acetone following the procedures of Donnelly et
al.(9). Additional rainwater was periodically added to the soils to
maintain a moisture content near field capacity.
  Soil samples of approximately 500 g each were collected from
each box approximately 24 hrs after each rainfall event. The sample
was obtained by compositing 6 to 10 randomly selected  soil cores,
each representing the entire 17 cm soil profile.
  Soybeans (Glycine max) were planted  in the control  and  waste
treated soils. The plants were harvested  upon appearance of the
first true leaves and extracted using diethyl ether. Soil and plant
samples were extracted using the method described by  Brown, et
al.8.  All soil and plant samples were stored at 0ฐC until extraction.
  Soil and waste samples were extracted with dichloromethane.
Twenty-five grams of the waste or waste-soil mixture were blended
with  six volumes of dichloromethane in a Waring Laboratory
blender at 9,000 revs/min for 30 sec. This extraction was repeated
twice or until the extracting solvent remained colorless. Solvent
extractions  were then  combined  and  taken to  dryness  on a
Brinkman-Bucci rotary evaporator. The  residue from this extrac-
tion was partitioned using liquid-liquid extraction into  acid, base
and  neutral fractions, following the procedures described by Don-
nelly et al.9.
  The Salmonella/microsome assay of Ames, et al.1 was used to
monitor the mutagenic activity of soil, plant and runoff extracts.
The Salmonella strains were supplied by  Dr. B. N. Ames (Univer-
sity  of California, Berkeley). The  procedural methods were the
same as Ames, et al.1 used except the overnight cultures were pre-
pared by inoculation into 10 ml of Nutrient Broth #2 (DC Biolo-
gical, Inc., Lenexa, KS) and incubated with shaking for 16 hr at
34 ฐC. Extracts were tested on duplicate plates in two independent
experiments in the standard plate incorporation assay  at a mini-
mum of 4 dose levels of the sample with and without enzyme acti-
vation (0.3 ml rat liver/ml S-9 mix) using strain TA98. [(Aroclor
1254-induced rat  liver  was  obtained   from  Litton  Bionetics
(Charleston, SC)]. All reagents and extracts were tested for sterility;
DMSO was used as a negative control  and 2-nitrofluorene and
benzo(a-pyrene were used as positive controls.
  Compounds present in the solvent extract of the runoff water
or the acid, base and neutral fractions of selected soil/waste samples
were tentatively identified with a Finnigan  OWA automated gas
chromatograph/mass spectrometer (GC/MS/DS) equipped with
a J&W Scientific (Orangeville, CA) fused silica capillary column
DB-5-30W. The DB-5-30W column had a liquid phase that was
bonded 1% vinyl/5% phenylmethyl polysiloxane. One-microliter
aliquots were used with a helium carrier gas flow of 36 cm/sec.
The gas chromatograph (GC) oven temperature program was 60 ฐC
for 1 min and then increased at 6 ฐC/min intervals to 260 ฐC with
a hold time of 12 min. The OWA unit had a splitless mode injec-
tor.  The software had a mass spectra library of 31,331 organic
compounds.
RESULTS
  Runoff water was collected from the Heswood and Bastrop soils
amended with the wood-preserving waste over a 1200-day period.
The mutagenic activity of the residual organic compounds  in the
runoff water from the Weswood soil increased with each subse-
quent sampling date. However, there was a slight decrease, both
with and without metabolic activation, in the number of revertants
induced by the sample solvent extracted from the run-off water
collected 1200 days following waste .application (Fig. 2).
          Weswood,  -S9
           1000
            10
              0.0     0.2     0.4     0.6     0.8     1.0
                         Dose/Plate (1 x ttf  kg)
          Weswood,  +S9
           lOOOl
•*• Control
-*- Dayl
•+• Day 180
-*• Day 360
•* Day 540
-o- Day 1200
— 2XBG
                            0.4     0.6
                         Dose/Plate (1 x
                           Figure 2
         Mutagenic Activity (TA98) Without (-S9) and With
                 + S9) Metabolic Activation for the
      Extracts of Runoff Water from Wood Preserving (PENT S)
                 Waste Amended Weswood Soils.
   Based on the mutagenic activity ratios given in Table 1 and
 illustrated in Figure 3, the mutagenic potential with activation of
 both the base fraction and the run-off water extracted from the
 PENT S-amended Weswood soil exhibited a consistent increase
 through 360 days following waste application. Without metabolic
 activation, the base fraction collected 1200 days following waste
 application exhibited a maximum mutagenic activity ratio of 9.7,
 and the acid fraction also exhibited the maximum value  of 2.8.
   According to Commoner10, the probability of correctly identi-
 fying a presumptive carcinogen increases as the mutagenic activity
 ratio (MAR) increases, or the probability of correctly identifying
 a carcinogen as a bacterial mutagen is greater than 99, if the chem-
 ical induces a MAR of greater than 2.5. Therefore, based on the
 statistical evaluation of Commoner10, there is a high probability
                                                                                        SAMPLING AND MONITORING     67

-------
                            Table 1
     Mutagenlc Potential (TA98) and Mutagenlc Activity Ratios of
     Residue Extracted from Soil and Runoff Water Collected from
      Wood Preserving Waste (PENT S) Amended Weswood Soil.


                        Specific Activity"

                   (Net Revertants/0.5  I  l(rฐ kg)
 Sample
None Historical
DHSO
B(a)p
2NF 1
Weswood 1
Control
waste 1



Soil/
waste 1



ISO



360



540



1200







Soil
Runoff
Crude
Acid
Base
Neutral

Acid
Bale
Neutral
Runoff
Acid
Base
Neutral
Runoff
Acid
Base
Neutral
Runoff
Acid
Base
Neutral
Runoff
Acid
Base
Neutral
Runoff
30 i a
26 ฑ 6
2 1 7
2171 ฑ 445
32 ฑ 8
33 i 8
6 i 3
LB-"
LB
LB

LB
LB
LB
LB
7 i 19
24 i 35
100 • 68
> i 10 <
9 i 5 <
77 i 17
1 ฑ 1 <
43 i 18
9 ฑ 6 <
47 i 40
14 i 6 <
163 s. 103
74 ฑ 36
253 ฑ 88
18 i 16 <
125 i 37
40 i 15
37 i 10
57J i 152
Not Tested
.2 67 i 16 l.S
.3 47 i 13 .3
141 i 17 .8
70 i 17 .9
126 i 11 .4
90 i 36 .4

104 i 11 .1
76 i 31 .1
75 ฑ 2>
42 i 9 .1
108 1 61 .9
267 i 115 .9
.8 184 i 41
51 i 32 .4
120 i 1ป 3.2
378 s. 22 10.2
124 i 17 3.4
.7 133 i 61 3.6
72 i 13 1.9
.8 341 ฑ 133 9.2
97 ฑ 23 2.6
.3 334 i 202 9
2. a 225 1 66 6.1
9.7 395 i 201 10.7
1 83 i 61 2.2
4.8 83 i 40 2.2
       Net revertants represents value obtained after revertants
       obtained from solvent  (DMSO)  alone have been subtracted from
       the total number of  revertanti for the sample.  The mean and
       standard deviation represents the average of three  runoff
       water extracts tested  on duplicate plates in two independent
       experiments (N-12).  -S9:  without metabolic activation. ปS9:
       with metabolic activation.
       Negative Control:  None ซ no  additions to top agar.
                        DMSO • 1 x 10-'L dimethyl sulfoxide.
       Positive Controls: B(a)p • Benzo(a)pyrene 5 x lO-'kg/plate.
                        2NF - 2-Nitroflourene 2.5 x 10-'"kg/plate.
       HAR is the nutagenic activity ratio obtained by dividing the
       net revertants by the  historical average for the DHSO solvent.
       LB represents incidence when  the number of revertant colonies
       was less than the background  induced  by the solvent alone.
                     180   360   540
                              Time (Dayซ)
                                                  1200
                           Figure 3
         Mutagenic Activity Ratio of Soil Base Fraction and
          Runoff Water Extract of PENT S Weswood Soil.
that there are genotoxic constituents in the runoff water 1200 days
following waste application. The detection of direct-acting muta-
gens in the runoff water and the impact of reducing the concen-
tration of selected organic compounds on the mutagenic potential
of the organics remaining  in the mixture is  difficult to predict

68    SAMPLING AND MONITORING
because mixtures of polycyclic aromatic hydrocarbons have been
shown to cause synergistic, additive and antagonistic responses"-
12
  The Weswood soil amended with the wood-preserving waste
(PENT S) exhibited an increase in mutagenic potential with meta-
bolic activation of the residual organics in the runoff water collected
540 days after waste application. Again, the base fraction of the
sample solvent extracted from the waste, amended soil collected
540 days following waste application gave a mutagenic activity ratio
(MAR) of 9.2, while the runoff water residue solvent extracted from
the same sampling period had a MAR of 9. The MAR of the runoff
water collected 1200 days following waste  application decreased
to 2.2, while the MAR of the  base fraction extracted  from the
waste-amended soil increased to 10.7.
  Compounds tentatively identified in the residue solvent extracted
from the wood-preserving waste-amended Weswood soil and runoff
water collected immediately after application included a broad
range of polycyclic aromatic hydrocarbons. The compounds iden-
tified in the fractions of the waste-amended soil collected immedi-
ately  after and 360 days following waste application have been
reported by Donnelly, et aJ.13. The mutagenic compounds (methyl
naplhalene, dihydroacenaphthalene, fluoranthene and pyrene) were
detected in the extract from the waste and waste-amended soil
initially and in the residue extracted from the runoff water imme-
diately following waste application.
  Biological and chemical analysis of the residue solvent extracted
from  the Bastrop soil  amended with  the wood-preserving waste
has been reported by Donnelly, et al.13. The mutagenic potential
with and without metabolic activation of the residual organics in
the runoff water from the wood preserving waste-amended Bas-
trop soil (Figs. 4 and 5) is similar to the Weswood soil but with
less dramatic increases over time.
  The chemical analysis data for the waste and residue  extracted
from  the waste amended soil have been summarized for  a com-
parison to  the compounds identified  in the runoff water (Table
2). The polycyclic aromatic compounds, methyl napthalene, methyl
dibenzothiophene, pyrene, and fluoranthene were tentatively iden-
tified in the residue solvent extracted from the waste and also from
the waste-amended soil extract collected immediately following
waste application. In the residue extracted from the run-off water
collected 360 days following waste application, fluoranthene and
pyrene were detected. These compounds were detected in  the waste
extract and the waste amended soil extract collected 360 days fol-
lowing application. The polycyclic aromatic compounds,  methyl
napthalene, dibenzothiophene, pyrene and fluoranthene were ten-
tatively identified in the fractions of the residue solvent  extracted
from the waste-amended soil sample collected immediately  follow-
ing waste application.  The compounds fluoranthene and  pyrene
were  also  detected in the  residue solvent extracted  from the
soil/waste mixture collected  360 days following waste application.
  The chemical analysis of the runoff water shows that methyl fluo-
rene, not detected in the waste but detected in  the waste-amended
soil samples, was also detected in the residue solvent extracted from
the runoff water collected from the Bastrop soil amended with the
wood-preserving waste immediately following waste application.
Hoffman,  et al.14 found that the non-methylated  fluorene could
be easily desorbed off sediments during the  filtration process.
Fluoranthene and pyrene, which  are  both mutagenic  and
cocarcinogenic2- "• 'ซ•l7, also were detected in the residue solvent
extracted from the runoff water collected 360 days following waste
application (Table 2).
  According to Bulman, et al.18, pyrene and fluoranthene seem
to be degraded  at comparable rates  due to similarities  in their
molecular  weights and physical properties. This may be one pos-
sible explanation for the increase in the number of net revertants
exhibited in the residue solvent extracted from the runoff water
collected 360 and 540 days following waste application with meta-
bolic activation.
  The residue extracted from the runoff water from the COMBO

-------
      Bastrop,  -S9
      1000
       ioo:
        10
•*•  Control
-*•  Dayl
•*•  Day 180
•*•  Day 360
•*•  Day 540
-O-  Day 1200
—  2XBG
          0.0     0.2      0.4      0.6      0.8

                       Dose/Plate (1 x TO*  kg)
                                               1.0
     Bastrop, +S9
      1000
Control
Dayl
Day 180
Day 360
Day 540
Day 1200
2XBG
                 0.2     0.4     0.6      0.8

                      Dose/Plate (1 x It6  kg)
                                               1.0
                       Figure 4
   Mutagenic Activity (TA98) Without (-S9) and With
            - S9) Metabolic Activation for the
Extracts of Runoff Water from Wood Preserving (PENT S)
             Waste Amended Bastrop Soils.
        45

        40-
        35-
        30

        ป
        20-

        15:

        10:
        5-
        0
               ISO
                    360   540
                        Time (Days)
                                             1200
                       Figure 5
Mutagenic Activity Ratio of Soil Base Fraction and Runoff
         Water Extract of PENT S Bastrop Soil.
                                                    Table 2
                        List of compounds tentatively identified in residue extracted from runoff
                        water from wood-preserving waste amended Bastrop soil, immediately
                                 following and 360 days following waste application.
                                                                                         GENETIC ACTIVITY *
                                                                                                             Haste "
                                                                                                                       Soil
                                                                      BASTROP TIME 1
Methyl naphthalene
Dimethyl napthalene
01 hydroacenapthal ene
Olbenzofuran
Trlmethyl naphthalene
Methyl flourene
Pentachlorophenol
Methyl dlbenzothlophene
Fluoranthene
Pyrene
4 Unidentified Peaks
BASTROP TIME 360
Fluoranthene
Pyrene
HI Mutagenic
HO Nonmutagenl c
Cl Carcinogenic
CC Cocarclnogenlc
CO Noncarclnogenlc
H1;CO
MO
Ml
X
MO
X
MO; CO
HO
Ml; Cl; CC
Ml; Cl; CC


Ml; Cl; CC
HI; Cl; CC
X
A
B
N

B
A, N
A. B, H
A. B. N
A. N
X
A
X
A, N
A, N


A, N
A, N
Unidentified
Identified In
Identified In
Identified In

B. N
A. B; N
K
A. B, N
A. B. N
N
A
B. N
A
A. B. N


A, B, N
A! B! N
In Any Fraction
Add Fraction
Base Fraction
Neutral Fraction

                             •    References for genetic activity Include: 15. 16, I 17.
                             **   References: 8 & 20.
                             **•   References: 13 I 20.

                         waste-amended Heswood soil induced an increasing number of net
                         revertants  through the  360-day sampling period with  a slight
                         decrease 1000 days following  waste application, both with  and
                         without metabolic activation. The data from the samples solvent
                         extracted from the soil samples collected  immediately following
                         each waste application have been reported19. The acid and base
                         fractions yielded the largest number of net revertants 360 days
                         following waste application (Figure 6). While chemical analysis was
                         unable to identify any of the constituents in the runoff water, over
                         half of the compounds identified in the waste20 were substituted
                         napthalenes.
                                                                      Soil
                                                                      Runoff
                                                   360
                                                    Time (Days)
                                                                         1000
                                                    Figure 6
                             Mutagenic Activity Ratio of Soil Base Fraction and Runoff
                                     Water Extract of COMBO Weswood Soil

                         The compound dimethyl  phenanthrene could be  a potential
                       source of mutagenic activity. Several substituted phenanthrenes
                       have been found to be mutagenic in the So/mone//a/microsome
                       assay with metabolic activation21. While the sample collected 1000
                       days following waste application induced a mutagenic response
                       both with  and without activation, the response was appreciably
                       lower than that of the sample collected 360 days following waste
                       application.
                                                                                       SAMPLING AND MONITORING     69

-------
  The Bastrop soil amended with the COMBO waste followed the
same general trend as seen in the Weswood soil although not as
pronounced (Fig. 7). Soil incorporation of the waste decreased the
mutagenic activity of the waste initially, but the mutagenic activity
in the extracts of the soil/waste and run-off samples was increased
over time. The increase in mutagenic activity observed over time
may have been a result of partial degradation of non-mutagenic
constituents, increased activity of metabolites of degradation or
a greater expression of synergism  between  waste constituents.
Interestingly, the maximum activity in the residue extracted from
the runoff water  was observed  180 days following waste appli-
cation, with a subsequent decrease  360 and 1000 days following
waste application with metabolic activation.
             301
                                          -ป Soil
                                          -•" Runoff
                                   Wuwood SoU, -S9
                                   Weiwood + COMBO, -S9
                                   Weiwood Soil. +S9
                                   Wuwaod + COMBO. +S9
                                                1000
                           Figure 7
         Mutagenic Activity Ratio of Soil Base Fraction and
           Runoff Water Extract of COMBO Bastrop Soil

   Expected chemical constituents in the COMBO waste include
 nitrogen containing heterocyclic hydrocarbons which have been
 shown to exhibit  high levels of mutagenic  activity in the Sal-
 monella/micTOSome assay22'a- These compounds may have been
 present in  concentrations below the  detection capabik'ty of the
 analytical instrument. Chemical analysis of the residue extracted
 from the runoff water collected immediately  after waste applica-
 tion detected five  substituted alkenes and 11 unidentified  com-
 pounds. The substituted alkenes also were identified in the fractions
 extracted from  the waste amended soils collected immediately
 following the initial rainfall event. The residue solvent extracted
 from the runoff water collected 360 days following waste appli-
 cation contained 12 compounds, and  all were at concentrations
 below the detection capabilities of the analytical instrument.
   Several attempts were made to grow bermuda grass or soybeans
 on the wood preserving waste amended soils. However, the toxicity
 of the PENT S amended soil was sufficient to inhibit the germina-
 tion of soybean seeds and elicit a phytotoxic response in bermuda
 grass sprigs. Conversely, the yield of plants grown on the refinery
 waste COMBO amended soil was greater than that for plants grown
 on control soils. In addition, the extract of  soybeans grown on
 refinery  waste amended soils appeared to contain low levels of
 bacterial mutagens (Fig.  8). The results indicate that the plants
 grown on waste amended soils induced a higher mutagenic response
 than those grown on control soils. Thus, in hazardous waste con-
 taminated environments,  food chain contamination may occur as
 a result of translocation into plants of organic mutagens from the
 soil.
   In summary, the differences in degradation, retention, mobili-
 zation and mutagenic activity of the extracted samples from the
 waste-amended soils and runoff water vary depending on the soil
 texture,  organic carbon  content and  past  soil-management
 practices. While each situation is site specific for hazardous waste-
 contaminated soils, biological analysis of soil and water samples
 may greatly enhance the results from chemical analysis in deter-
 mining recalcitrance and  migration  of  hazardous constituents
            20
              01     234567
                     Dose (gram equivalents per plate)

                           Figure 8
   Mutagenic Potential of Diethyl Ether Extract of Plants Grown on
             COMBO Waste Amended Weswood Soil

capable of causing genetic damage from contaminated sites.

CONCLUSIONS
  The use of short-term bioassays to measure the release of genc-
toxic constituents from hazardous waste-contaminated soils has
been shown to be an asset in circumstances where the results of
chemical analysis  are  inadequate  and  inconclusive.  Chemical
analysis  alone  fails  to account for  synergistic, additive  and
antagonistic reactions which may  occur when compounds are
introduced into a biological system.
  Additionally, physical and chemical characterization may pro-
vide information regarding the runoff potential,  binding  and
absorptive capabilities of the soil affected with hazardous consti-
tuents. The bacterial mutagenicity of the residual organics extracted
from the runoff water collected from the waste amended Weswood
soils was greater than the response observed for the runoff water
from the waste amended Bastrop soil.  This result may have been
due to differences in texture, organic carbon content and past
management practices of the soil. While chemical analysis could
not identify the chemical compounds, due to their low concentra-
tions or masking in the complex matrix of the mixture, the bio-
logical analysis  detected possible mutagenic compounds. Results
from a chemical analysis must be extrapolated to estimate the toxi-
cological end point in a biological system. When assisted by chemi-
cal analysis,   the  detection   of  mutagenic  compounds  in
environmental samples using bioassays may support the determi-
nation of health risks associated with  exposure to contaminants
migrating from soils contaminated by hazardous wastes.
  Mutagenic compounds in a soil/waste mixture may be retained
in the soil, degraded over time, translocated in plants or removed
by runoff water. Results from this study indicate that detoxifica-
tion, degradation and  removal of mutagenic constituents from
hazardous  waste-contaminated  soils may require three years or
longer. Furthermore,  the runoff water may require additional treat-
ment before it is considered non-hazardous. Finally, these data in-
dicate that mutagenic compounds may be introduced into the food
chain through translocation from contaminated soil into  plants.

ACKNOWLEDGEMENT
  Research was funded by the U.S. EPA's Office of Research and
Development, Cincinnati, Ohio. Chemical analysis was performed
by Dr. Don Kampbell, a Research Chemist,  at the U.S. EPA's
Robert  S.  Kerr Environmental Research Laboratory in Ada,
Oklahoma.

REFERENCES
  1. Ames, B.N., McCann, J.  and Yamasaki, E., "Methods for Detecting
    Carcinogens and Mutagens With the So/mone//a/Mammalian Micro-
    some Mutagenicity Test," Mutat. Res.,  31,  1975, 347-364.
70     SAMPLING AND MONITORING

-------
2. McCann, J., Choi, R., Yamasaki E. and Ames, B.N., "Detection of
   Carcinogens as Mutagens in the Salmonella/Microsome Test: Assay of
   300 Chemicals." Proc. Nat. Acad. Sci., 72, 1975, 5135-5139.
3. Rinkus, S.J. and Legator M.S.. "Chemical Characterization of 465
   Known or Suspected Carcinogens and Their Correlation With Muta-
   genic  Activity in the  Salmonella  Typhimurium  System," Cancer
   Research, 1979, 3289-3318.
4. Claxton, L.D., Halsh, D. and Stead, A., "Analysis, Including Corre-
   lation With  Carcinogenicity of the Salmonella Typhimurium Bioassay
   Data Bases  by Chemical Class," Environ.  Mutag., 8,  1986,  18.
5. Tennant, R.H.,  Stasiewicz, S. and Spalding, J.W.. "Comparison of
   Multiple Parameters of Rodent Carcinogenicity and In Vitro Genetic
   Toxicity," Environ.  Mutag.,  8, 1986, 205-228.
6. Ames, B.N., "Identifying Environmental Chemicals Causing Mutations
   and Cancer," Science, 204, 1979, 587-593.
7. Brusick, D.J., Principles of Genetic Toxicology. Plenum Press. Inc.,
   New York,  NY, 1980.
8. Brown, K.W., Donnelly, K.C. and Thomas J.C., "Use of Short-Term
   Bioassay to Evaluate Environmental Impact of Land Treatment of
   Hazardous Waste,"  "Final Report to Robert S. Kerr Environmental
   Research Laboratory." U.S. EPA, Ada, OK, Grant No. R-807701-01,
   EPA 600/2/84/135:386, 1984.
9. Donnelly K.C., Brown. K.W. and Scott B., "The Use of Short-Term
   Bioassays to Monitor the Environmental Impact of Land Treatment of
   Hazardous Wastes," In (Water, M.D., S. Sandhu, J. Lewtas, L. Clax-
   ton, N. Chernoff, and S. Nesnow, Eds., The Application of Short Term
   Bioassays in the Fractionation and Analysis of Complex Environmen-
   tal Mixtures HI, Plenum Press, New York, NY. 1983.  58-78.
10. Commoner, B.,"Reliability of Bacterial Mutagenesis Techniques Car-
   cinogenic and Non-carcinogenic Chemicals," U.S. EPA, Office of
   Research and Development,  Washington,  DC, EPA-600/1/76/022,
   1976.
11. Hass, B.S.,  Brooks, E.E., Schumann, K.E. and Dornfeld, S.S., "Syner-
   gistic, Additive, and Antagonistic Mutagenic Responses to Binary Mix-
   tures of Benzo(a)pyrene and Benzo(e)pyrene as Detected by Strains
   TA98 and  TA1OO in the Sa/mone//a/Microsome Assay," Environ.
   Mutat., 3,  159-166, 1981.
12. Haugen, D.A. and Peak, M.J., "Mixture of Polycyclic Aromatic Com-
    pounds Inhibit Mutagenesis in Sa/mone/fa/Microsome Assay by Inhi-
    bition of Metabolic Activation, " Mutat. Res., 116, 1983, 257-269.
13.  Donnely, K.C., Davol, P., Brown, K.W., Esteri, M. and Thomas, J.C.,
    ' 'Mutagenic Activity of Two Soils Amended With a Wood Preserving
    Waste," Environ. Sci. Tech., 21, 1987, 57-64.
14.  Hoffman, E.J., Mills, G.L., Latimer, J.S. and Quinn, J.G., "Urban
    Runoff as a Source of Polycyclic Aromatic Hydrocarbons to Coastal
    Waters," Environ. Sci. Tech., 18, 1984, 580-587.
15.  Kaden, D.A., Kites, R.A. and Thilly, W.G., "Mutagenicity of Soot
    and Associated Polycyclic Aromatic Hydrocarbons to Salmonella
    Typhimurium," Cancer Res., 39, 1979, 4152-4159.
16.  Goldschmidt, B.M., In J.M. Fontag, Ed,  Carcinogens In Industry and
    the Environment, Dekker, New  York,  NY, 1981, 283-343.
17.  DHEW, "Survey of Compounds Which Have Been Tested for Car-
    cinogenic Activity," Department of Health, Education and Welfare,
    United States Public Health Service Publication 149, 1968-1969 Vol.,
    1972; 1961-1967 Vol., Sect.  I and II,  1973;  1970-1971 Vol. 1974;
    1972-1973 Vol., 1975; Suppl. 1, 1967; Suppl. 2, 1969, Washington, DC,
    United States Government Printing Office, 1969.
18.  Bulman, T.L., Lesage, S., Fowlie, P.J.A. and Webber, M.D., "The
    Persistence of Polynuclear Aromatic Hydrocarbons in Soil," Environ-
    ment  Canada, Environmental  Protection Service,   Waste-water
    Technology Center, Burlington, Ontario, 1985, PACE Report No. 85-2
19.  Brown, K.W., Donnelly, K.C., Thomas, J.C., Dabol, P. and Scott
    B.R., "Mutagenic Activity of Soils Amended With Two Refinery
    Wastes,"  Water, Air and Soil Pollut.,  29, 1986,  1-13.
20.  Donnely, K.C., Brown, K.W., Thomas, J.C., Dabol, P., Scott, B.R.
    and Kampbell, "Evaluation of the Hazardous Characteristics of Two
    Petroleum Wastes," Hazardous Waste and Hazardous Mat., 2, 1985,
    191-208.
21.  La Voie, E.J., Tulley-Freiler, L., Bedenko, V. and Hoffman, D.,
    'Mutagenicity of  Substituted Phenanthrenes  in Salmonella Typhi-
    murium." Mutat. Res., 115, 1981, 91-102.
22.  Rosenkranz, H.S.  and Mermelstein,  R., "Mutagenicity and Geno-
    toxicity of Nito-Arenes: All Nitro-Containing Chemicals Were Not
    Created Equal," Mutat. Res., 114, 1983, 217-267.
23.  Donnely, K.C., Jones, D.H. and Safe, S.H., "The Bacterial Mutage
    nicity of Nitropolychlorinated Diabenzo-p-dioxins." Mutat. Res., 169,
    17-22.
                                                                                               SAMPLING AND MONITORING     71

-------
                     Application of  Data Quality  Objectives  at  a
                          Superfund  Remedial  Investigation  at  a
                                    Former  Municipal  Landfill

                                                   Marc P. Lieber
                                                 ICF  Incorporated
                                                 Washington, DC
                                                   John D. Frost
                                               ICF Technology, Inc.
                                             Pittsburgh,  Pennsylvania
                                                    Paula  M.  Lia
                                     U.S. Environmental  Protection  Agency
                                     Waste  Management  Division,  Region I
                                              Boston, Massachusetts
                                             Michael  Amdurer, Ph.D.
                                               Ebasco Services, Inc.
                                                Arlington, Virginia
ABSTRACT
  The U.S. EPA recently issued guidance on using Data Quality
Objectives for Superfund remedial  response  activities.  This
guidance contains new requirements for the process of formulating
the scope of Superfund remedial investigations,  planning site
sampling efforts and specifying analytical laboratory requirements.
The requirements are new and detailed and affect almost  all
remedial investigation planning activities. This paper presents an
example of how data quality objectives were used successfully in
planning a remedial investigation at the Old Springfield Landfill
Superfund site.

INTRODUCTION
  One  of the key analytical problems facing  investigators of
hazardous waste sites under Superfund  is how to design a site in-
vestigation program that is both efficient and comprehensive.  On
one hand, investigators attempt to collect sufficient data to present
clearly the contaminants, routes of migration and exposure and
areal extent and volume of contamination so that cleanup alter-
natives can be evaluated with precision.
  On  the other hand, sampling efforts, laboratory analyses,
installation of monitoring wells and other common investigation
techniques can be costly. The average cost of a RI/FS ranges from
$0.6 million to over $1  million and requires 1 to 2 years to plan
and complete. In order to save  time and money, investigators
constantly attempt to find new ways to streamline site investiga-
tions while ensuring the collection of adequate amounts of data.
  The U.S. EPA issued guidance entitled "Data Quality Objec-
tives for Remedial Response Activities" (EPA  540/6-87/003A,
OSWER Directive 9335.0-7B March, 1987) to assist Superfund data
users in developing site-specific data quality objectives (DQOs).
The stated purpose of the DQO process is to "help assure that data
of sufficient quality are obtained to support remedial response
decisions, reduce overall costs of data sampling and analysis
activities and accelerate project planning and implementation."

ROLE OF DQOs IN SUPERFUND INVESTIGATIONS
  DQOs serve two primary functions in Superfund investigations.
First, they specify the quality of data required to support U.S. EPA
decisions during remedial response activities. If a DQO is specified
for each data collection activity conducted at a Superfund site, then
both the investigator and data user or decisionmaker are in agree-
ment on the decisions to be made and on the data expected to
support  each decision. This agreement  is the second primary
function of DQOs: they permit explicit review of the rationale for
and execution of investigation activities in support of the key
decision  to be made at Superfund sites.
  For example, in designing the sampling and analysis program
to characterize groundwater contamination at a Superfund site,
an investigator must make many specific choices to include well
numbers, locations, materials, etc. Additionally, the investigator
must determine if rigorous analytical and quality control procedures
of the U.S. EPA's CLP are required or if less rigorous analysis
and QC  (e.g., standard commercial laboratory analysis by U.S.
EPA methods, or a mobile laboratory) would suffice.
  The choice of sampling materials, methods and analytical pro-
cedures is made by determining the decision that will be made based
on the data produced. The DQO process is the procedure for
defining: (1) the decision, (2) the quality of data required to support
the decision and (3) the specific sampling and analytical techniques
required to produce this level of data quality.
  The Superfund DQO process, although containing specific steps
and being a relatively new requirement, is not considered a wholly
new or unexpected requirement by those familiar with U.S. EPA
procedures or the Superfund program. The origin of DQOs may
be found in two areas. First, in May, 1984, the U.S. EPA Deputy
Administrator requested development of DQOs for all U.S. EPA
programs. This request was followed in October, 1984 by a checklist
for DQO review  that  included such clearly useful steps as iden-
tifying the decision makers, stating the decisions to be made based
on the data, specifying the questions addressed by the data collec-
tion effort and explaining how the data collected will provide the
precision and accuracy necessary to answer these questions.
  Additionally,  during the  summer of 1984,  U.S. EPA's CLP
experienced a shortage of capacity that resulted in postponement
or cancellation of selected field sampling efforts and lengthened
turnaround times for receipt of analytical laboratory data. U.S.
EPA initiated several steps to increase capacity and manage demand
for CLP analyses. One such demand management effort was to
encourage Superfund site investigations  to screen samples using
field instruments such as portable gas  chromatographs, or to
increase use of non-CLP laboratories to reduce demand on CLP
laboratories. This effort required investigators to review critically
the objectives of each sample collected and analysis requested and
72    SAMPLING AND MONITORING

-------
to determine which samples could be  eliminated,  screened or
diverted to laboratories outside of CLP. The DQO process was
developed, in part, to ensure that the resources required for each
sample analyzed were justified as necessary to a successful Super-
fund investigation.

USE OF DATA QUALITY OBJECTIVES AT THE
OLD SPRINGFIELD LANDFILL SUPERFUND SITE
  Although the U.S. EPA's guidance on DQOs was issued in
April, 1987, U.S. EPA had developed this guidance over a 2-year
period. The U.S. EPA's DQO Task Force included representatives
from U.S. EPA Headquarters, Regional offices and remedial con-
tractors. The substance of the DQO guidance, if not the documen-
tation, was available to participants in the U.S. EPA's Superfund
remedial response program though draft versions of the DQO
guidance and reports of Task Force discussions.
  The U.S. EPA's REM contractor, Ebasco Services, Inc., incor-
porated DQOs as a standard procedure to be used in determining
the scope of RI/FS beginning in Fall,  1985. DQOs were used
explicitly in scoping approximately 30 remedial work assignments,
including 20 RI/FS.
  This paper presents a case study showing how DQOs were used
successfully in scoping the Superfund RI/FS at the Old Spring-
field Landfill Superfund Site under the REM program. There are
two purposes of the case study: first, to illustrate how DQOs were
an effective analytical and planning tool for determining the scope
of work for a RI/FS, and second, to indicate practical experience
with DQOs that can be used by other site investigators to amplify
and emphasize some key issues that are not addressed in the DQO
guidance document.

Old Springfield Landfill RI/FS: Site Description
  The Old Springfield Landfill Superfund site is located in the
Town of Springfield, Vermont, approximately 1 mile southeast of
the commercial and residential center of the town. The 27-acre site
was a landfill which accepted municipal and industrial waste from
local machine shops from  947 to 968. The relatively flat site is
situated on an upland terrace with slopes that descend steeply to
the local drainage to the north, east and west. The site is situated
100 ft above Seavers Brook to the west and the Black River to the
north and east.
  Following closure of the landfill the  site  was developed as a
mobile home  park. The park contains 38  mobile homes and
approximately 78 residents. To the north of the landfill is a con-
dominium complex. Thirteen permanent homes are located on the
road adjacent to the mobile home park. Residents of the park
receive township water, but wastes go to a septic/leach field system.
The steep, eastern side slopes  often slump  or slide during wet
seasons.
  Available information indicates that the landfill accepted indus-
trial waste from local machine tool and metal plating industries
into existing ravines and trenches dug in the sand. Bulk and
contaminated waste oils, cutting oils, solvents, sludges, heat treating
salts and acid plating and etching wastes were reportedly disposed
of at the  site. Disposal of the wastes has resulted in contaminated
seeps emanating from the steep side slopes, contaminated soils and
sediments, contaminated groundwater and the potential contami-
nation of the receiving streams.
  Previous investigations have included various studies by State
agencies  and an RI by the U.S. EPA FIT contractor in 1984. The
U.S. EPA continued with supplemental investigations and analyses
through  1986, when it initiated the subject REM III RI/FS.

DQO Process: Stage 1
  The broad outline of the U.S. EPA's DQO development process
is depicted in Figure 1. Following the identification of data users
and a thorough evaluation of the existing data collected by previous
investigators, Stage 1 of the process involves developing a concep-
tual model of the site and specifying objectives for the project.
                         STAGE  1
                 IDENTIFY DECISION TYPES

                • IDENTIFY & INVOLVE DATA USERS

                • EVALUATE AVAILABLE DATA

                • DEVaOP CONCEPTUAL MODEL

                • SPECIFY OBJECTIVES/DECISIONS
                      STAGE 2
             IDENTIFY DATA  USES/NEEDS

             • IDENTIFY DATA USES

             • IDENTIFY DATA TYPES

             • IDENTIFY DATA QUALITY NEEDS

             • IDENTIFY DATA QUANTITY NEEDS

             • EVALUATE SAMPLINGVANALYSIS OPTIONS

             • REVIEW PARCC PARAMETERS
                          1
                         STAGE 3
           DESIGN  DATA COLLECTION PROGRAM

           • ASSEMBLE DATA COLLECTION COMPONENTS

           • DEVELOP DATA COLLECTION DOCUMENTATION
                         Figure 1
                  DQO Three-Stage Process
  In Stage 1, the RI/FS team attempted to identify the key issues
that  the U.S. EPA must address in conducting the Superfund
cleanup. As discussed in SARA, the investigator must evaluate the
mobility, toxicity and volume (MTV) of site contamination. The
following is a summary of the issues addressed in Stage 1.

What are the present and future risks at the site?
  Previous investigations indicated that hazardous leachate was
emanating from the steep side slopes of the landfill. Direct contact
or ingestion of sediment and leachate could present a potentially
significant health threat, but because of limited access and the steep,
rough terrain,  it was  determined that the potential  threat  of
exposure was low to minimal. Based on this evaluation, local
authorities installed a fence around the west side seep area to limit
access. Future  risks include leachate transported to surface and
groundwater. Surface  soil  contamination was not  considered a
significant threat, but subsurface contaminants could be encoun-
tered if shallow excavations were dug. The subsurface sources of
contamination are also believed to be contributing to groundwater
contamination. Sufficient data did not exist to determine whether
groundwater contamination presents a present or future risk.
Finally, potential risks  may be posed through contaminant release
to the air and transport of contaminants to surface water through
discharge from groundwater.
                                                                                     SAMPLING AND MONITORING    73

-------
                Objective
          I. Providr More Conclusive
             Evaluation of Preaent Risk
                                                                                  Table 1
                                                       Summary of Data Needs for (he Old Springfield RI/FS
                                               Date Gap Element  or  factor
                                                    IQ ^ Evaluated
 A. Risks resulting from
    landfill off gain
 I. Risks rasulting trim
    dlract contact with
    subsurface materiels
C. Verification of preaent
   rlaks in domestic Mils

0. Extent and trend of
   contamination in laachate
                                                                                    Data Heeded
                               1.  Buantltlatlva data  on VOC
                                  content  of  air
                               I.  varIf leal ion of  raportad
                                  Mastc  disposal areas
                                                                            Proposed Data
                                                                            Cathedra />ctlปlttes
  Air monitoring and  Quantitative  sampling  .
  analysis
  Geophysical  characterlutlon/ecreenlng
  ualng EH 51  and (ft
                                                                             1. Eitanslva sampling and analvsis   -Subsurface sampl Ing and analysis  using
                                                                                of subeurfece ปซites and soils     boreholes xltti teat pit verification
                                                                              I. Quantitative dels on MSL
                                                                                contaminants in domestic Mils
                                                                              t. Location of sll contaminated
                                                                     •Sailing of domestic Mils for CLP
                                                                      analysis

                                                                      Seep napping and seซp saapl ing and
                                                                      analysis
          II. Assess Migratory Potential
              and Future Impact of Site
              contaminants
          III. Evaluation of Source
              Control Technologies
         IV. Evaluation of Marwjgement
             of Migration Technologies
 A. Contaminant source
    characteristics for fete
    and transport model ing
 I. Significance of ahalloii
    ftoป (future risks el
    seapa)
                                               C. Mydrogaologic condmone
                                                 governing transport in
                                                 till and shallow bedrock
                                                 (future health risks to
                                                 ground vater users and
                                                 future iapacts to Seavers
                                                 Irook and Hack liver)
                                              0. Mechanisms leading to
                                                 contamination in
                                                 Seavers Irook area
 A.  Capping, excavation and
    dispoaal or  treatment
                                              R. Subaurface containment  or
                                                 diversion of  ground water

                                             C. Septic syatem replacmaant
D. Relocation of residents

E. no action alternative

A. Seep remediation
   (fencing, collect and
   dispoae, collect and
   treat,  septic tank
   replacement)
                                             R. Ground water controls
                                                (pump and  treat)
 1 Hazardous Substance* list
                                                                             2. Quantitative del* on contsml
                                                                                nants In seep water and
                                                                                sediments
                                                                             llama IRI and 112
                                                                             1. Location, extant,  elevation
                                                                                and hydraulic characteristics
                                                                                of shallow aquifer
                                                                             2. Quantitative date  on
                                                                                contaminants In ehallow
                                                                                ground Hater
                                                                             }. flevellon of ell aejor springs
                                                                                and seeps

                                                                             1. Items IR1 and 112
                              2. Additional on stratigraptiy
                                 and structure

                              J. Additional on horixontal
                                 gradients

                              4. Vertical gradients

                              5. Hydraulic conductivity in till
                                 and bedrock
                              6. CEC. IOC. porosity
                                 of aquifers
                              7. pM of  ground water

                              1. Item IRI

                              2. Items  Mil  3

                              J. Item  IICI  6

                              4. Evaluation of  each transport
                                 mechanism proposed in
                                 Section 2.0

                              1. Extent,  volume,  and chemical
                                 characteristics  of  sutcurface
                                 waste  dim IRI  and lซ7>

                              I. Itw  HIM
                              1. Revised water balance baaed
                                on water consumption

                              2. Hydrogeology of shellow MOM
                                (llama Mil-))

                              ). Uaate locations (Item IRI 2)

                              5. All  items  In I

                              6. All  items  in I and  II

                              1. Location/extent of  seeps  and
                                direct contact haiard (Itam
                                101-2)

                              2. Seep flow  ratea

                              1. Influence  of shallow flow
                                (Mam 1111  3)

                              t. Pilot treetment study

                              1  Contaminant migration pathways,
                                extent of  ground water
                                contamination and hydraulic
                                characteristics (all llama  in
                                no
                                                                   •Inelsl lซticn.  testing,
                                                                    15 shallow ptciomitar*
                                                                                                                                         eampl ing of up te
                                                                    •Survey elevation of  all aajor
                                                                     breakout  points

                                                                    •Installation testing and  aampl ing  of
                                                                     7 till  Malls (6  nested for  vertical
                                                                     gradient  determination)

                                                                     Same as C.I
                                                                                                                  Same as C.I


                                                                                                                  •Testing and confirmatory sampling of all
                                                                                                                  uistlnj Mils
                                                                                                                  Installation and teat ing of up lo five
                                                                                                                  bedrock Mils
                                                                                                                  •Collect soil  images for analysis
                                                                     See  stove

                                                                     Set  above

                                                                    •See  above
                                                                     See stove
                                                                                                                 •Rctriev*  ueter useage  recorda
 See above

•Sea above

 See above



 See above

•See above


•larch scale tests*

 Sea above
74     SAMPLING AND MONITORING

-------
Where are the sources of contamination?
  It is impossible  to  evaluate  future risks from groundwater,
surface water and subsurface soil without knowing the location
of contaminant sources within the landfill. Once located, investi-
gators can evaluate the volume of source material, model the
mobility of the contaminants, evaluate the toxicity and analy/e the
site characteristics and receptors with respect to potential remedial
measures.

What Potential existing risks to health and the environment should
be characterized?
  At least three risks may exist for which indicative data have not
been collected These include air, groundwater and subsurface soils
The air beneath and around the trailers must be sampled to deter-
mine if unsafe levels of methane or elevated concentrations of
volatile organic contaminants exist in the breathing space of resi-
dents. The aquifer feeding the seeps and the one used for a potable
water source by residents adjacent to the site must be further evalu-
ated Finally, the uncertainty of the location and depth  of buried
waste as related  to shallow  excavations presents an undefined
potential exposure  risk.

What are the contaminants of concern?
  Thirteen contaminants of concern have been selected based on
the frequency and  distribution  within all media, concentrations
detected and the toxicological properties of the contaminants. The
contaminants of concern  include: six volatile organics  (benzene,
1,2-transdichloroethene, toluene, trichloroethene, vinyl chloride
and xylene);  three semi-volatiles  (bis-2-ethylhexyl phthalate,
benzo(a)pyrene  and fluoranthene);  three  inorganics  (barium,
chromium and cyanide);  and PCBs.
  Based on the conceptual model developed for the site and the
key issues involved, the contractor met  with the U.S. EPA to
develop goals and overall objectives of the investigation in order
to develop alternatives for remediation. With the key objective of
remedial alternative identification and ROD signature in mind the
discussions were focussed  on the steps needed to reach these goals.
These discussions led  to Stages 2 and 3 of the  DQO process.
                                  DQO Process: Stage 2 and 3
                                    Stage 2 involves identifying the data types, quantity and quality
                                  needed and selecting sampling and analytical approaches to support
                                  the objectives identified in Stage 1. The U.S. EPA's DQO guidance
                                  offers the following five analytical levels:

                                  • Level I —field screening or analysis using portable instrumen-
                                               tation  such  as photoionizers  or  organic vapor
                                               analyzers
                                  • Level II —field analysis using more sophisticated portable ana-
                                               lytical  instruments or mobile laboratories
                                  • Level III—analysis performed by off-site laboratories using U.S.
                                               EPA-approved procedures
                                  • Level  IV-CLP routine analytical services
                                  • Level  V-nonstandard analysis

                                    Table 1 presents a partial summary of data needs at Old Spring-
                                  field to characterize risks and assess future migration of hazardous
                                  substances.  The information from this summary exhibit can be
                                  translated easily into  a specific set of sampling and analysis options.
                                  For example, the purpose of air monitoring is to quantify data on
                                  volatile organic compound content of air for use in assessing risk.
                                  This use indicates the need for a sampling plan and analytical
                                  protocol offering sufficient precision and accuracy to use in a risk
                                  assessment. In contrast the soil TOC and CEC data are used in
                                  fate and transport modeling and therefore Level III is acceptable.
                                  Screening for VOAs, PCBs and CN in the mobile laboratory is
                                  used as a contaminant indicator for followup data quantification
                                  and therefore Level II is acceptable.
                                    Stage 3 of the DQO process involves selection of sampling and
                                  analysis options and design of the site investigation plan. Table 2
                                  summarizes the DQO analytical levels  used at Old Springfield.
                                  Because almost all analyses were performed to characterize risks
                                  to human health and the environment,  DQO Level IV  (CLP
                                  analysis) was  selected in most cases.
                                    The notable exception was the source characterization activity.
                                  The goal of this task was to compile conflicting data from anecdotal
                                  accounts of site activity, soil gas and geophysical investigations
Phase
I
I
I
I
Media
Air
Air
Residential
Well
Soil
Borings
                                                               Table 2
                                         Old Springfield Landfill Site Data Quality Objectives and
                                                      Type of Analysis by Media
                                   Number of
                                   Samples           Duplicates     Blanks
                    Test Pits
                    Soil

                    Leachate

                    Sediment
                    Ground Water
           1  or II
             II
                     Ground Water
                                     8                    1

                                     9                    1
                                    100% (collected)
                                    25% (screened)         6
                                     6% (CLP)              4
                                     5 (REM III  lab)       1
  12                    2

  12                    2
   5  (shallow)           1
  18  (existing)
   7  (till)             2
   2  (bedrock)
5-10  (shallow)
   3  (intermediate)
 All  samples

   4  (bedrock)**         1
                                                                      1
                                          POO Level

                                          I

                                          IV

                                          IV
                                          I
                                          II
                                          IV
                                          III

                                          IV
IV

IV
IV
IV
IV
IV
IV
IV
I

IV
FID screen for total  volatiles

HSL VGA

HSL organic
HSL inorganic (CN)

Visual,  pH, Conductivity, OVA
VGA*,  PCB  (field screening)
HSL organics, HSL Inorganics (CN)
TOC, Cation Exchange  capacity

HSL organic
HSL inorganics (CN)

HSL organic and inorganic (CN)

HSL organic and inorganic (CN)
HSL organic and inorganic (CN)
HSL organic and inorganic (CN)
HSL organic and inorganic (CN)
HSL organic and inorganic (CN)
HSL organic and inorganic (CN)
HSL organic and inorganic (CN)
Field pH,  conductivity,  temperature

HSL organic and inorganic (CN)
            •Screening Analysis to include: 1,1-dichloroethane,  tetrachloroethene,  trichloroethene, chloroform,
                                    t-l,2-dichloroethene,  2-butanone,  benzene, toluene, ethylbenzene, xylene,  1,1,1-trichloroethane
           ** optional

           DQO Levels:   1   Field Instrument Screening, II    Mobile Lab Analysis,  III   REH III  Laboratory,  IV   CLP Analysis
                                                                                          SAMPLING AND MONITORING     75

-------
and aerial photography interpretation and to establish an on-site
grid  for sampling and locating the  source areas The  program
required the collection of soil samples  from 100 15-ft deep test
borings from across the site. Several  factors indicated that Level
IV analysis was not desirable or necessary. First, over 100 samples
would be collected and many samples would be clean resulting in
wasted money for Level IV CLP analysis.  Second, the analytical
method had to be capable of very quick turnaround analysis to
allow on-site decisions to continue drilling vertically or increase
the number of borings to characterize the vertical and horizontal
extent of contamination. To do this, a series of field screening and
mobile laboratory steps was required. Those samples found to be
contaminated using DQO Level I screening in the field triggered
additional drilling and were sent to the on-site mobile laboratory.
Samples found to be contaminated in the mobile laboratory (DQO
Level II) were sent to CLP for  Level IV  analysis so the data could
be used later in the evaluation of risk. Because the data were used
to determine additional sampling locations, rather than to assess
risk, it was determined that a mobile laboratory (DQO Level) would
be  most  appropriate to deliver accurate  results  and  quick
turnaround.

PRACTICAL CONSIDERATIONS  IN USING
THE DQO PROCESS
  The Old Springfield Landfill RI/FS was one  of the first Super-
fund RI/FS to be planned and implemented with explicit use of
DQOs. As such,  it offers several valuable practical insights into
using DQOs  at Superfund sites.

  Definition  of DQO Levels
  The U.S. EPA's DQO guidance specifies five DQO levels. From
a practical viewpoint, the levels related to specific sources of labora-
tory analysis rather than levels  of data quality. Moreover, the five
DQO  levels overlap significantly.  For  example,  a portable gas
chromatograph could be considered DQO Level 1 or 2; the quality
of data produced in  a mobile  laboratory could be equivalent to
that produced in an off-site laboratory, yet could be classified as
DQO Level 2 or 3. DQO Level 5, non-standard analysis, can range
from quick-turnaround CLP-SAS analysis to a specialized screen
to engineering or treatability analysis to a parts-per-trillion GC/MS
or MS-MS analysis.
  From a practical viewpoint, RI/FS planners may consider three
basic data uses: screening, engineering and confirmational analyses.
Within each data  use, many practical considerations will indicate
the specific source of laboratory analysis. For example, although
the data quality of a mobile laboratory may be equivalent to an
off-site laboratory, the mobile laboratory may only be cost-effective
for large numbers of samples for which quick  turnaround is re-
quired (as was the case at Old Springfield). Table 3 presents a sum-
mary of the basic DQO levels cross-referenced with the DQO levels
in the  U.S. EPA  guidance.

                           Tibia 3
                    Summary of DQO Level*
Level I
(Hซซlth and Safety)
Level II
(Site Characterization)
Level V
(Special Bice screening)
ENGINEERING

LEVEL II
(Feadblllty Study)
CONFIRMATIONAL

Level IV
(Sice Characterization)
(Rllk Asaeiinenc)
(Litigation)

Level V
Level III
(Site Characterization)  (High-reiolutlon,
(Deilgn)               Matrix Interferen
                     or other epeclel
Level V                protocol!)
(Special engineering
 analyse*)
                                              At Old Springfield, considerable discussion occurred over which
                                              of these three DQO levels to select. Eventually, DQO Level 4 (CLP
                                              analysis) was  selected for two primary reasons: first, because it
                                              offered data of known quality and second, because the cost of
                                              analysis was funded by a separate U.S. EPA CLP account rather
                                              than from  the budget set aside  for the RI/FS. Given  these
                                              considerations, it is likely that many other RI/FS planners would
                                              make a similar choice, even possibly if the Level 3 or 5 analysis
                                              were less costly than  the Level 4 analysis.

                                              PARCC Considerations in DQO Levels 2, 3 and  4
                                                Ideally according to the U.S. EPA DQO Guidance, the end use
                                              of the data collected  during the RI should  define the necessary
                                              PARCC parameters, (i.e Precision, Accuracy, Representativeness,
                                              Completeness and Comparability). During the work plan develop-
                                              ment,  the precision, accuracy and completeness goals would be
                                              established and used to select the sampling and analysis methods.
                                              This investigation, like all field investigations was quite different
                                              from the ideal situation. The goal of the waste area delineation
                                              task was to analyze (qualitatively) as many samples as possible as
                                              quickly as possible so that decisions could be made in the field con-
                                              cerning additional boring locations, monitoring well locations and
                                              quantitative analysis needs. Although  PARCC parameters were
                                              examined briefly, the  need  for quick analysis of large numbers of
                                              samples was the primary consideration in selecting a DQO Level
                                              2 mobile laboratory rather than a fixed laboratory.
                                                A review of the Level 3 and 4 historical precision and accuracy
                                              data showed a level above that required of our data use, therefore
                                              Level 2 was selected. Because historical precision and accuracy data
                                              are not established for this particular sampling and field analy-
                                              tical procedure, the PARCC decision was more an intuitive one
                                              based on experience at other sites with our laboratory. As shown
                                              in Table 4, precision and accuracy data were unavailable from the
                                              DQO guidance.
                                                                        Table 4
                                                            Precision and Accuracy Information
                                               l.n
                                               Tt
                                               T.
                                               To
                                               El
                                               ซ7
                                               I,
                                               PCI
                                                  Compound
                             hloroซthซna
                             •chloroathcn
                                                                    l*rซ5 tit toa   	     	Accuracy
                                                              l^vvl':  Uvซl )  Uyซl %     Lvvcl 2   Uvvl )
 Cost Considerations in DQOs 3 and 4
   The U.S. EPA's DQO guidance indicates that DQO levels 3,
 4 and 5 may be selected for data to be used in the risk assessment.
  In the case of the Old Springfield RI, a reduced level of pre-
cision and accuracy was acceptable because the "level of concern"
was the detection limit of the mobile laboratory instrumentation.
Any detectable finding alerted the lab technician to package the
sample for CLP (Level 4) analysis so the quantitative data could
be used in  the risk assessment. In previous investigations performed
by ICF for the U.S. EPA's Superfund program, the detection limit
could not be used as the "level of concern." At a site in Louisiana
contaminated  with hexylchlorobenzene, the contaminant was a
known  carcinogen at a concentration below  the 20 jig/1 CLP
detection level. Therefore, special consideration of the PARCC
parameters was needed to evaluate the  sampling and analysis
options. ICF used special wells, sampling procedures and sample
concentrations to obtain analysis to  the 9 pg/1 value required.
  An additional practical consideration is related to the use of
PARCC parameters relative to the evaluation of the soil matrix.
Soils are not homogenous and contaminant attenuation in soils
of varied horizons, parent material and locations within a site will
vary significantly. Obtaining sample duplicates for evaluation of
precision when collecting split-spoon samples is extremely difficult
and varies  from boring to boring based on percent sample recovery.
Mixing to collect representative samples is not recommended when
collecting samples for volatile organic analysis. For the evaluation
76    SAMPLING AND MONITORING

-------
of accuracy, the debate continues over what represents a soils blank.
  In conclusion, unlike water samples, many factors come into
play when using PARCC parameters to evaluate the soil matrix.
When using the PARCC information to develop a sampling and
analysis program, it is very useful to review the information with
the data users and to consult your chemist in regard to the analyt-
ical equipment capabilities and sample volume needs and the sam-
pler  who understands the difficulties in obtaining  each type of
sample. If all involved are in agreement as to what samples need
to be collected, what volume needs to be collected, what data will
be obtained and how the data are to be used, the sampling trip,
evaluation of the data and report preparation will be efficient and
effective.

CONCLUSION
  The DQO process is a welcome development for improving the
efficiency of RI/FS planning and the efficacy of site investigation
efforts. By following the DQO process, investigators can ensure
that sampling and analytical data serve the anticipated uses. It is
anticipated that, as DQOs are used widely, practical experience
will lead to improvements in the guidance, training and quality
of implementation.
ACKNOWLEDGEMENT
  The  authors wish to thank  Robert Fellman of the Ebasco
Services, Inc. REM III Zone Program Management Office and
Richard McCracken of ICF Technology for their assistance in
planning the sampling  effort at the Old Springfield Landfill  site
and for their guidance in incorporating DQOs into the REM III
RI/FS  planning process.
                                                                                     SAMPLING AND MONITORING    77

-------
                         Remediation  at  a Major  Superfund  Site
                        Western  Processing  — Kent,  Washington
                                              Kenneth A. Lepic, P.E.
                                             HDR Infrastructure,  Inc.
                                                Seattle,  Washington
                                                   Allan R.  Foster
                                                      ERT, Inc.
                                              Redmond, Washington
ABSTRACT
  Western Processing is a former industrial waste processing and
recycling facility and major  superfund  site  located in  Kent,
Washington. Investigations conducted between 1982 and early 1986
have identified 73 contaminants found in soil samples including
20 metals and 53 organics and 46 priority pollutants, including both
metals and organics, in groundwater samples.
  An extensive and unique Held program was developed and
implemented in the fall of 1986 to evaluate the nature and extent
of on- and off-site contaminants to design remediation measures.
Magnetometer/gradiometer,  electromagnetic induction and ground
penetrating radar were  utilized.  EM terrain  conductivity and
inphase component maps, total magnetic Held and vertical gradient
maps and radar target maps were subsequently developed. A com-
  PUGET SOUND
                                                  WATER
                           Figure 1
                         Vicinity Map
posite of these maps supplied the location and characterization of
subsurface anomalies. Geophysical results were successfully used
to guide the soil sampling program.
  An on-site  laboratory was equipped  with two gas
chromatographs, a high-pressure  liquid chromatograph and an
atomic absorption spectrophotometer to screen for 26 different
chemical parameters in each  sample collected. The parameters list
included total cyanide, metals, volatile organics and semi-volatile
organics. Turnaround time was 24 hr., and over 1,500 samples were
analyzed. In addition, several screening methods were developed
specifically for contaminants encountered at the site. Positive agree-
ment was obtained when the screening results were compared to
the results reported by  the off-site laboratories.

Introduction
  The Western Processing property is located a few miles south
of Seattle, Washington (Figure 1).  It was operated as an industrial
waste processing and recycling facility from 1961 to 1983. During
its period of operation. Western Processing handled, processed and
recycled animal by-products, brewers yeast and a wide variety of
industrial waste products including solvents,  flue dust, battery
chips, acids and cyanide solutions. In 1982, the U.S. EPA placed
the  property on the NPL  and issued an administrative order
requiring the cessation of all operations and the initiation  of site
remediation.
  In April 1983, the U.S. EPA initiated an emergency response
and sampling program for the removal of drums and impounded
liquids from the site. Later in 1983, the Washington Department
of Ecology (Ecology) completed a project to control storm water
runon and runoff. This project included stockpiling contaminated
sediments, grading and paving portions of the site and constructing
of berms. These emergency responses were followed by a 1984
surface cleanup by  certain PRPs.
  Final remedial action, which  will first address subsurface con-
tamination, began  in mid-1987. This private party action  will be
one of the largest to date in the United States. The investigation
described in this paper was conducted to provide additional physical
and chemical information specific to the site and immediate vicinity
for  development of certain  aspects of the final remedial action.

STUDY OBJECTIVES
  During August 1986, HDR Infrastructure,Inc., initiated a study
of the Western Processing site and adjacent properties (Figure 2).
The objectives of the study were to:
• Sample and analyze soil and waste materials to develop chemical
  data with which priorities for removal of contaminated materials
78    SAMPLING AND MONITORING

-------
  or other remedial techniques could be evaluated
• Develop correlations between visual appearance of soil and waste
  materials and contamination within the site
• Determine the locations of buried drums, tanks, utilities and
  process lines within the site and as these utilities exit into adjacent
  property
• Determine the extent of surface and subsurface contamination
  on property bordering the site

GEOPHYSICAL SURVEY METHODOLOGY AND RESULTS
  The specific objectives of the geophysical survey were to locate
buried drums, tanks, utilities and process lines and determine the
locations of all abandoned or active utilities or process lines leaving
the site  and crossing into  bordering property.
  EM and MAG data were collected on 10-ft. centers, and ground
penetrating  radar (GPR)  profiles were obtained along parallel
tracklines on 20-ft. centers. Additional GPR lines were run in other
orientations and spacings  where  required  to gather more site-
specific data. Two additional GPR survey tracklines were run on
10-ft. centers inside and outside the perimeter fence in order to
locate and map subsurface utility corridors or process lines which
exited or entered the site. A single traverse of EM and MAG data
was collected as a parallel offset adjacent to the perimeter fence.
240
                  240
     SCALE  1' 240'
                            Figure 2
                     Western Processing Site
                        Kent, Washington
                       Investigation Areas

Geophysical Systems
  Geophysical systems deployed included: an EDA OMNI IV total
field magnetometer with internal digital memory and a base station
magnetometer; a GEONICS EM-31 with digital recorder; and a
GSSI  120 mHz GPR with a digital tape recorder.  All geophysical
systems were operated in the field according to technical procedures
developed specifically for this program.
  The GEONICS EM-31  operates on the principles of electro-
magnetic induction. A transmitter coil radiates an electromagnetic
field signal which induces eddy currents in conductive materials
in the earth beneath the instrument. These eddy currents in turn
produce  a secondary magnetic  field which is  detected by the
instruments receiver coil. The EM-31 measures the terrain con-
ductivity by comparing  the strength of the quadrature  phase of
the secondary field to that of the primary (transmitter) field. The
quadrature phase signal arrives at the detector coil out-of-phase
(i.e. in quadrature) with respect to the primary field. This  measure-
ment represents a weighted average of the electrical conductivity
of  the soil to a  depth of approximately 20 feet beneath the
instrument.
  The terrain conductivity data on the Western Processing site were
recorded  in  the  field  using  an   OMNIDATA  digital
POLYCORDER. In addition to providing an efficient data inter-
face with a personal computer, this recording technique permitted
the simultaneous acquisition of two distinct channels of  informa-
tion from the EM-31, the "quadrature phase channel" and the
"inphase channel." The quadrature phase channel yields the soil's
apparent conductivity in millimhos/m, while the inphase channel
provides a measure of the terrain magnetic susceptibility, given in
parts per thousand of the primary field. Experiments by GEONICS,
the manufacturer of the  EM-31, indicate that the inphase channel
is about twice as sensitive to the presence of buried metallic objects
as the quadrature phase. Thus, the inphase  channel of the EM-31
constitutes a very effective metal detector.
  The EDA OMNI IV Total Field Magnetometer with Gradiometer
measures the disturbance in the earth's total field created by iron
and steel (ferro-magnetic) objects. The magnetometer is particularly
effective in the search mode since the high permanent magneti-
zation present in iron pipes and stacks of drums leads to large local
disturbances in the  earth's magnetic field.
  In addition to measuring variations  in the earth's total field
strength,  the  EDA OMNI IV magnetometer simultaneously
measures the vertical gradient of the earth's total magnetic  field.
Vertical gradient data tend to have greater lateral resolution than
total field data. Therefore, multiple buried objects which appear
as a single composite total field anomaly often can be identified
individually on vertical gradient maps. This  greater spatial resolu-
tion is important on hazardous waste sites in reducing the  area and
effort of search at individual anomalies.
  The magnetometer data were corrected for diurnal variations
and contoured in total magnetic field and vertical gradient format
for  analysis and interpretation. The vertical gradient data were used
to define the locations of magnetic anomalies which were generally
expressed on the total magnetic  field presentation.
  GPR operates on exactly the same basis as aircraft radar. The
instrument transmits a pulse of electromagnetic energy downward
into the subsurface at a frequency of 50 kHz. This signal is reflected
by  differing layers  of soil or by buried objects. The  density,
moisture content and composition of the subsurface soils affect
the strength of  the signal reflected by subsurface  interfaces.
Therefore, the instrument receives signals at different times and
amplitudes from the soil layers and presents a continuous cross-
sectional geophysical profile of the subsurface features along a
given trackline. The ground penetrating radar system was utilized
in conjunction with the magnetometer and EM-31 systems to
characterize shallow subsurface anomalies.  The radar system was
used to refine and extend the data from the other systems and to
locate and track buried  utilities.

Geophysical Survey Results
  EM terrain conductivity and inphase component maps, total
magnetic field and vertical gradient maps, and a radar target map
were developed. A composite of these maps and reference to his-
torical documentations identifying past land use activity on the site
allowed the characterization of anomalous subsurface conditions
A segment of the composite EM, MAG and GPR anomaly map

                    SAMPLING AND MONITORING    79

-------
is shown in Figure 3.
                                     —~     ^s to*   \
                                     M&
                                    3    •  ^  ^c<
             100              0               100
                            FEET
                           Figure 3
              Composite Geophysical Anomaly Map

  Thirty-nine anomalous regions were identified. A brief narrative
of the nature of each anomalous region including an estimate of
the possible nature of the anomaly was prepared to assist in the
subsurface exploration and soil testing portion of this  investigation.
A map showing a summary of the terrain conductivity (EM) data
also was  prepared. Areas of terrain conductivity  greater than
100 mmhos/m are considered anomalous. Large areas of the site
were determined to fall within this category. These areas of high
terrain conductivity may represent shallow layers of soil saturated
with groundwater of high specific conductance or locations of
buried conductive wastes and debris. An example of the usefulness
of the EM-31 terrain conductivity is depicted in Figure 4 where
a signature  of a  buried  utility  trench is identified. By further
illustration,  the radar signature of a buried 6-in.  steel pipe which
exited the site is presented in Figure 5.
  The results of the geophysical interpretations were relied upon
to focus the investigation of  subsurface conditions.  These results
were used successfully to locate, excavate and sample a wide variety
of subsurface features including buried drums, subsurface tanks,
buried utility corridors and  process lines.

SUBSURFACE INVESTIGATIONS
Drums, Tanks and Utilities
  A map showing some of the drums, tanks, utilities and process
lines identified during the investigation is presented in Figure 6.
Exposure of these objects was accomplished by trackhoe equipped
with an explosion shield.
  Two buried tanks were located and sampled during the investi-
                                                                                             Figure 4
                                                                        Terrain Conductivity Signature of a Buried Utility Corridor
gation. The first tank consisted of a three chambered unit which
was accessible at ground surface via two exposed manways. Liquid
in the tank was decanted to allow sampling personnel to enter the
manways in supplied air/protective clothing and obtain samples
of sediments from each chamber and from a 10-in. inlet pipe. The
second structure consisted of a buried metal process tank. Repre-
sentative samples were obtained  from the bottom of the tank.
  Numerous buried pipelines were located and exposed. Samples
of pipe contents and the bedding material surrounding the pipe
were obtained using special non-sparking brass tools or the trackhoe
bucket.

SoU  Borings
  As shown in Figure 7, 207 soil borings and 92 test exploration
sites were completed  and soil  samples collected  for chemical
analyses. The soil borings were completed using a B-61 and/or B-23
Mobile Auger drill rig with an 8 in. (outside diameter) hollow-stem
auger. The borings generally ranged from 8-ft. to 30-ft. in depth
below ground surface.
  Borings typically were completed to 15 ft. with sample intervals
at 3 ft. above 10 ft. and continuous below 10 ft. In some instances,
additional samples at greater depths were obtained to evaluate and
characterize the vertical  extent  of contamination.  Additional
sampling was discontinued once visual appearance of the sample
and  field screening using an organic vapor analyzer (OVA)
indicated  levels  within  acceptable limits  at  that depth. Upon
completion, the borehole was backfilled utilizing the tremie method.
A 50% bentonite/50% cement slurry mix  was  pumped downhole
from total depth to ground surface.
80    SAMPLING AND MONITORING

-------
                                                                         g        6        d        o        3        o
                                                                         ?        o        S        o        o        o
                                                                         2        i-        i        ฃ        i        f.

                                                                           H-H-H-+-H--H-|-M-f-| - - T--4-++++
 X
 o
      1C
      1*
                                                  (MOUND WMFACt
                          22 FKT
                          Figure 5
               Ground Penetrating Radar Profile
              Illustrating the Geophysical Signature
                Of a Buried 6-in. Diameter Pipe
Test Explorations
  A total of 92 test sites were excavated on a 75-ft. center grid
pattern using a backhoe. Test sites were excavated to a depth of
12 ft.) providing sidewall stability could be maintained (or to the
point were groundwater was encountered. Groundwater was en-
countered at shallow depths ranging from about 3 ft. to 12 ft. be-
low ground surface. The length of each exploration was nominally
10 ft., but some extended to 35 ft. The locations of a number of
test sites are shown in Figure 7.
  Soil samples also were extracted using a 36-in. (3 in. diameter)
Shelby tube sampler which was affixed to the backhoe bucket. Once
the test sites were completed and photographed, each excavation
was restored using material initially removed from the excavation.
The  grossly contaminated materials,  sludges or other materials
registering strong  OVA (or similar  instrument) readings were
segregated and placed in the upper portion of the excavation un-
exposed to the surface.
ENVIRONMENTAL MONITORING PROGRAM
  An air monitoring program was implemented to ensure adequate
protection of both the field team and the surrounding community.
Work area monitoring was conducted to identify action levels where
personnel protection levels must be upgraded. Continuous upwind
and  downwind perimeter monitoring was conducted.
                                                                                 — CENTER LINE BICYCLE PATH
                                                                                                                       AREA  X
                                                                                             Figure 6
                                                                            Segment of Drum/Tank/Utility Investigation Map
  Direct reading, real time instruments were used to determine total
gases and vapors, cyanide, gamma radiation and combustible gas;
particulate concentrations also were measured. The field instru-
mentation used at Western Processing included:
• OVA 128: total organic vapors
• HNU PI 101: total organic vapors
• Handheld aerosol monitor (HAM): total particulates
• Gastech CGI: combustible gases
• Ludlum 19:  gamma Radiation
• Monitox Compur 4100: cyanide
• Draegerpump and colorimetric detector tubes:  cyanide and
  methane
• Hi-volume air samplers: suspended particulates
• Recording meteorological station: wind direction and speed.
  Air monitoring  was regularly  conducted  at 16 fixed locations
around the site perimeter to detect any possible off-site migration
of airborne contaminants. In addition, monitoring  was conduct-
ed at each sample location to determine adequate protection levels
and  ensure  worker, safety.  These  monitoring  procedures  are
described below:

• Borehole and excavation site monitoring: an OVA, HNU and
  HAM were used to monitor the breathing zone at the drill rig
  and  backhoe during the subsurface exploration and sampling
  activities. A cyanide detector and combustible gas indicator were
  used on a regular basis.
• Drum, tank and utility monitoring:  an OVA and HNU, radia-
  tion  detector, combustible gas indicator  and cyanide monitor
  were used to test the atmosphere within containers for flammable
  vapors.
                                                                                       SAMPLING AND MONITORING    81

-------
	 BTl* r
\\
ฐ ซ•ซ • iWi ซ.
II^FICI U*ปM "*" I'
MtiHMICint WN" • kWCI 11
1MVU IKMM Mป* "J |
MI <•*.• ••*ซ)> • nnซ tป
TปI im io>'mป <*•ซ(•
ป * w<ซ ••,• M hBE*tป>
mซ* |i MKATtl 'ป ft aWM
S!t'
'"
/vJ
• J

* : -1,
•1*
1
/ "!'
-4"A
^
+




+.ป. .ซ,
i!
mf
ARC* VIII
.•ป
".-
1MQ •— < 1***
_SHS...-~ 	 ^ML.
c
AREA VII „
•n ^ j
a
& AT,
• -• "^T '
A.EJ!
C,i sH^

•0 5.
,-•" .ป>.••..ซ
, ., uu_
1+ 5-—, -
" s-'-
S, f'/0 _*
""*" ^.^-'
ฃA
+ \ A-*
', AA
.//
•,7 A 4.8
,
y.o tf .^
AREA IX
IK
A ฃ ฃ.5



,
A '" ••
^ *J*
. f~ T
i"' *ซ. A
xrw
^-i-,*
* 'S.
,
J A
•~; &
ป ! A
~ I s,
- __*,
"!'•
A & -9
•
M" 'a*
' ! A
•j"*
!"Ji
,11
. ^
'ii ซ•ซ

. *
• '"ป
, A
• : "; ,
* '"
rr^j- •"' ซ
i 7?. ซ

•"'li~ , i
-o* '"
- -A !
1' , •
• 4*
"^ •" i
'"" A
Ttf-A
''Mป.Q !
>** •*' 3
,,:K
.. ; ~r ^ 1
™T •
",+ -i i
-' !
C •"<: ;
- • - ,
tm •*"• '
?i^* "
                              i   i   :.   i  A  A
                                   AREA III-S
                                          ,,  i  ซ. i,
                           Figure 7
                   Investigation Layout Map
THE ANALYTICAL LABORATORIES
  The investigation program resulted in the collection and analysis
of 1515 samples of soil and non-soil materials. These samples were
forwarded for  laboratory  analysis.  The  analytical  program
developed for the project  included three laboratories.  An on-site
laboratory was set up to provide rapid screening of soil samples
for 26 chemical constituents. Two off -site laboratories provided
priority pollutant analyses on 272 samples. One off-site laboratory
                                                                    performed solid waste leach procedure (SWLP) testing on 127
                                                                    samples. These laboratories performed all analytical work in con-
                                                                    formance with a detailed quality assurance program developed
                                                                    specifically for this project.
                                                                      The on-site laboratory consisted of a sample processing facility
                                                                    and an analytical facility. This design was selected to reduce the
                                                                    possibility of sample contamination. Samples were delivered to the
                                                                    processing facility  for weighing, extraction and digestion. The
                                                                    resulting sample extracts were then transferred to the analytical
                                                                    facility for analysis. This facility contained two Hewlett-Packard
                                                                    Model  5809A gas  chromatographs (GC)  fitted with automatic
                                                                    sample and  data  processors;  a Waters high pressure  liquid
                                                                    chromatograph (HPLQ coupled to a WISP auto sampler, a Waters
                                                                    Model 490 multiwavelength detector, and a Perkin Elmer Model
                                                                    LS-4 uv-fluorescence spectrophotometer with automatic data pro-
                                                                    cessing by two Spectra Physics Model 4290 integrators; a Perkin
                                                                    Elmer Model 2300 atomic absorption (AA) spectrophotometer; and
                                                                    a Dionex Model 4002i ion chromatograph (1C). Except for the AA,
                                                                    these instruments were fully automated and ran continuously for
                                                                    the duration of the project. The on-site laboratory was designed
                                                                    to screen 20 samples per day for inorganic  and organic con-
                                                                    taminants, however daily processing rates  peaked at 40 samples.
                                                                    Samples were analyzed for the 26 chemical constituents listed in
                                                                    Figure 8. The constituents are grouped according to the analytical
                                                                    equipment used  by the on-site laboratory.
                                                                      In support of the investigation, the on-site laboratory provided
                                                                    analytical  results the day  following  sample  collection and, if
                                                                    required, a  1-1/2 hr. turnaround for any specific sample. Fast
                                                                    access to the results of chemical analyses enabled field coordinators
                                                                    to direct  field operations based on daily sampling reports. To
                                                                    achieve this rapid turnaround and still be able to process the desired
                                                                    number of samples, the on-site laboratory used screening analyses.
                                                                    Generally, these methods were simple modifications of existing U.S.
                                                                    EPA procedures. In some cases, new analytical screening methods
                                                                    were developed  specifically  for the  contaminants  at  Western
                                                                    Processing.
                                                                      Rigorous quality assurance testing was carried out to assure that
                                                                    the on-site testing provided reliable analytical results. The quality
                                                                    control program included blank and duplicate samples, analytical
                                                                    replicates, matrix spikes and reference materials. The blank and
                                                                                  ION
                                                                              CHROMATOGRAPHY

                                                                              Cadmium
                                                                              Chromium (hex)
                                                                              Chromium (total)
                                                                              Copper
                                                                              Cyanide (total)
                                                                              Lead
                                                                              Nickel
                                                                              Zinc
               HPLC

        Oxazolldone              *
        2,4-d1chlorophenol       f
        Phenol
        2,4-dimethyl phenol
        b1s(Z-ethylhexyl)phthalate
        Isopherone
                                  HEAOSPACE ANALYSIS
                                    CHROMATOGRAPHY

                                  1,1-dlchloroethane
                                  Chloroform
                                  Ethylbenzene
                                  Methylene Chloride
                                  Tetrachloroethy1ene
                                  Trans-1,2-dichloroethene
                                  1,1,1-trlchloroethane
                                  Trlchloroethene
                                  Toluene
                                  1,1-dlchloroethene
                                                                                                             GAS
                                                                                                        CHROMATOGRAPHY
                                                                                                        PCB Aroclors (total)
                                                                                                        Benzo(a)pyrene
Legend:  * denotes analysis was performed on-site and off-site,
            (If blank, analysis was performed only on-site).
        I analysis was only performed on off-site properties.
                          Figure 8
                 Screening Parameters — Soil
                     On-Site Laboratory
                   Western Processing Site
82    SAMPLING AND MONITORING

-------
duplicate  samples represented  an additional 20% of the total
investigation samples.  Duplicate samples were presented to the
laboratory in a totally blind fashion. One screening reference
material, oxazolidone, is not readily available and therefore had
to be synthesized.
  As  with most intensive sampling and analysis projects, some
problems were encountered. The  fast turnaround allowed these
problems to be quickly identified and suitable modifications made.
This process  generally  took 1 to 2 days and prevented a halt to
field activities or the accumulation and possible loss of a large
number of samples. The analysis of samples for volatile organic
compounds  presents an example  of the rapid resolution of a
problem. The high vapor pressure of volatile organic compounds
makes their loss from contaminated soil a major concern. The field
sampler sealed the sample in a container with a Teflon septum and
metal hub directly at the point where the sample was collected.
This container then remained sealed throughout the analysis. The
sample was  analyzed by inserting a needle through the septum
sealing the container and extracting atmosphere from the head
space within the container U.S. (EPA Method 5020).
  A surrogate was added to each sample and was used as a gross
indication of any problems with the analysis. Occasionally, samples
showed a low recovery or a complete absence of the surrogate. This
loss was quickly traced to the field samplers who had not properly
cleaned the top of the  vial before  capping.  The fast turnaround
enabled the sampling team and on-site laboratory to modify their
procedures the following day to prevent recurrence of this problem.
  Quality assurance was a major consideration throughout all
phases of this project.  The on-site laboratory adhered to an
analytical quality assurance program designed specifically for this
project. The  investigation required that over 8,900 individual
chemical tests be run on-site. Additionally, 3,200 separate quality
assurance tests were run in support of this program. Direct com-
parison of the results from split samples run by the  on-site
laboratory and the off-site CLP  laboratories,  and from inter-
laboratory comparison  samples prepared by the on-site laboratory,
demonstrated good comparison between the screening and U.S.
EPA-CLP analyses. Examination of  the results for the blind
duplicate analyses demonstrated that the analytical precision was
excellent.
  The on-site laboratory quality assurance met all the requirements
set forth in the sampling and analytical plan. No significant con-
tamination was found in the blank samples. The standard curves
exhibited good fit on data points, and calibration data were within
20%  of original data.

Analytical Results
  The on-site laboratory results for all screening analyses and other
soil properties were reported in March 19871. Priority pollutant
analytical results and all other aspects of the study were reported
in June 19872.
  As noted  above, SWLP testing was performed on 127 of the
investigation samples.  SWLP test protocol includes a number of
cycles whereby samples are mixed with a leaching  solution then
filtered. The filter cake is retained and the filtrate is analyzed for
individual concentrations of six  metals.  In the subsequent cycle(s),
the retained filter cake is leached and the filtration/analysis steps
are repeated. For the Western Processing samples, two leaching
solutions were used—groundwater obtained from the vicinity of
the site and a 0.1 molar solution of sodium citrate. Four test cy-
cles of each leach solution were performed on every SWLP sam-
ple. Additional cycles of the two leaching solutions were performed
on selective samples. Taken together, 2180  cycles of SWLP test-
ing were conducted.
   In all, approximately 300,000  units of analytical data were gener-
ated. These data were entered into a custom designed relational
database management system developed in the R:BASE  System
V database  environment. Each of the  three laboratories entered
analytical data directly into menu-driven modules of the database
system which were designed and provided for their use. Analytical
results were reported to HDR on floppy discs or telephone modem
on a daily "screening" or weekly (priority pollutant/SWLP) basis.
Micro-computers located on-site were used to track sample infor-
mation including sample location, depth, chain-of-custody, etc.
The computers also were used to perform data reduction/data
sorting and report generation.
  In addition to daily reports of analytical results, the data were
reported in ad-hoc configurations as needed by the study team,
the client and representatives of the U.S. EPA and Ecology. For
instance, results of the previous day's results could be sorted and
reviewed to identify  locations where additional  samples were
needed. These systems were invaluable to a  fast-track intensive
investigation of this nature.
  During the investigation, the results of the screening analysis were
reviewed to determine appropriate locations for additional borings
and test exploration  sites. The computer data were sorted  into
desired  categories, and three dimensional contour plots of the
screening data were prepared. This information served both as a
tool in this decision making process and a method of providing
a graphical display of interim data to the U.S. EPA and Ecology.
Examples of the three-dimensional plots are shown in Figure 9 for
cadmium and  chromium.
                     CaoVnun Concentrations h Sol
                   Chromium Concentration in Soil

                           Figure 9

                      SAMPLING AND MONITORING    83

-------
PROSPECTIVE
   HDR Infrastructure, Inc. is a member of the multi-disciplinary,
multi-firm team selected to design, construct and operate the final
remediation program for the Western  Processing site.  The data
from the investigation discussed in this paper and from prior data
collected at the site are being used to complete the design of final
remediation. That program includes the following:

• Excavation, transport and off-site disposal of 10,000 yds.3 of
   grossly contaminated soil and/or buried wastes
• Design, construction  and operation of capture systems for
   shallow and intermediate depth groundwater
• Design, construction and operation of groundwater treatment
   facilities including air  stripping and chemical unit  processes
   (Figure 10)
• Installation of 60 long term stations for monitoring compliance
   with groundwater and  surface water standards
• Design and construction of closure facilities for the site, including
   a roller compacted concrete cap, and runoff/runon  controls.

CONCLUSION
Careful  delineation of  contaminant  plumes was achieved by
extensive geophysical surveys and designing a sampling  plan that
allowed  for  modifications  if hot spots were found.  A  rapid
turnaround of analytical results was obtained to direct Held oper-
ations and to allow for modifications in procedure or technique,
should problems be encountered.
   The final result was a very detailed delineation of the soil con-
tamination and other properties for use in the subsequent final site
remediation design and operation. The use of on-site screening of
the samples allowed the entire site investigation to be completed
                            Figure 10


in ninety days (less than half the time it took to complete the
272 U.S. EPA-CLP analyses) with assurance that the sampling was
complete and comprehensive.
REFERENCES
1.  "Analytical Screening Data Report, Western Processing Site, Kent,
   WA.," HDR Infrastructure, Inc., Mar. 1987.
2.  "Geophysical Survey and Site Sampling Investigation, Western Pro-
   cessing Site, Kent, WA.," HDR Infrastructure, Inc., June 1987.
84    SAMPLING AND MONITORING

-------
                            Low Level Volatile Organic  Analysis

                                                Michael S. Zachowski
                                                  Rutgers University
                                      Department of  Environmental Science
                                            New Brunswick, New Jersey
ABSTRACT
  The evaluation of uncontrolled hazardous waste sites requires
groundwater evaluation both on and off site. Monitoring wells are
located and constructed and samples are collected and subjected
to chemical analysis. The investigator hypothesized that conven-
tional chemical analysis may be inadequate to characterize ground-
water conditions at these sites. Therefore, samples were collected
at a CERCLA site and subjected to both U.S. EPA Invitation for
Bid Contract Laboratory Program analysis utilizing purge and trap
methodology as well as closed loop stripping (CLS)sample prepara-
tion followed by GC and GC/MS analysis. Data evaluation con-
cluded that CLS was a valid and important analytical tool necessary
to achieve detection limits 50  to 500 times lower than CLP con-
tract required  detection limits. Utilizing the CLS procedure,
monitoring wells contaminated at the pg/1 level could be identified
when conventional CLP analysis would report them as below the
method detection limit.

 Introduction
  Leaks,  spills and improper  land disposal of hazardous wastes
have created numerous uncontrolled hazardous waste sites which
have come under investigation by State and Federal agencies. A
large portion of the environmental evaluations conducted at these
sites focuses on the groundwater route of contaminant migration.
Monitoring wells are constructed in varying geologic strata at
distances  away from the site  in an attempt to characterize the
vertical and horizontal extent of groundwater contamination. With
standardized sampling  techniques  and established U.S. EPA
Invitation for Bid Contract Laboratory Program (U.S. EPA IFB
CLP) analytical protocols, any given monitoring well can be deter-
mined either to be contaminated or not contaminated at the /ig/1
level. Using this methodology, the contaminant plume is defined
and a groundwater remediation scheme proposed, if required.
  During the course of one such investigation, it was noted that
while no contamination was detected at the required CLP detec-
tion limits,  careful examination of the mass spectra provided by
the laboratory indicated that analyses were present at levels lower
than the contract required detection limits (CRDL) associated with
U.S. EPA  analytical methods. Therefore an alternate sample
preparation and analytical technique was evaluated as a possible
future protocol by which substantially lower detection limits in
groundwater samples might be achieved. If this technique could
be refined and validated, the capillary fringe of a contaminate
plume could be established, alerting investigators to potentially
deteriorating environmental conditions. Extended contaminant
plumes could  be identified and tracked prior to contamination
reaching concentrations detectable by conventional analysis. This
would avoid a potential public health threat posed by drawing
potable water from a contaminated aquifer.

PRINCIPLES OF PURGE AND TRAP
  Purge and trap concentration is used as a sample preparation
step prior  to  analysis  by gas chromotography (GC)  or  gas
chromotography/mass  spectrometry (GC/MS)  methods. The
technique was developed by the U.S. EPA for  the analysis of
volatile organic compounds in  water. These volatile constituents
are removed from the aqueous phase by bubbling an inert gas
through the sample. Once in the gas phase,  the compounds are
swept to an absorbent trap. At  the end of the purge step, the trap
is rapidly heated and backflushed to a chromatographic column.
In this  method, the sample is concentrated approximately one
1000-fold.

PRINCIPLES OF CLOSED LOOP
STRIPPING ANALYSIS
  Closed Loop Stripping Analysis (CLSA) was developed by Grob
and is similar  to the purge and trap procedures of Bellar and
Lichtenburg described above. In CLSA, solubilized volatile and
low molecular weight semi-volatile organic compounds are removed
from water by spraying with headspace gas saturated with water
vapor. Volatile and semi-volatile compounds are partitioned from
a 1-1 volume of water sample into the sample headspace; the par-
titioned compounds are then swept to a microparticulate activated
carbon trap. The activated charcoal retains organic compounds
while allowing the headspace gas to pass through. The headspace
gas is then  recycled to repurge the sample via an inert pumping
system. At the end of the purge cycle, the trap is removed and the
organic compounds are chemically desorbed with  a small amount
(25 ml) of solvent, typically carbon disulfide. An aliquot of this
solvent is then subjected to separation and analysis utilizing either
GC or GC/MS techniques. Due to the large number of compounds
concentrated in  a typical closed loop stripping analysis run,
capillary columns normally are  required  for high resolution
chromatographic separation.

EXPERIMENTAL DESIGN
  To compare the CLSA procedure vs. the conventional U.S. EPA
IFB  CLP analysis, the  following experiment was  designed. A
suitable CERCLA Site was identified and site conditions were
evaluated. This selected site was  studied and monitoring wells were
inventoried. A monitoring well was chosen that had previously been
sampled and subjected to CLP analysis with below method detec-
                                                                                  SAMPLING AND MONITORING    85

-------
tion limits (BMDL) found for volatile and semivolatile compounds.
This well was immediately adjacent to the site and downgradient
of the contaminant  plume,  but the well's screened interval was
beneath a clay stratum believed to be acting as an aquiclude. This
well was sampled utilizing appropriate methodology to exclude ex-
traneous contamination, and the samples were split between the
CLSA experiment and a U.S. EPA 1FB CLP certified laboratory.

SITE DESCRIPTION
  The subject site is an inactive landfill located in the New Jersey
Atlantic Coastal Plain physiographic province, where the soils are
primarily sandy and  highly permeable. These unconsolidated
sediments are known as  the Cohansey Formation, which is con-
sidered to be the most productive aquifer in New Jersey. The site
had operated as a sand and gravel excavation area and then as a
landfilling operation for municipal and industrial wastes.  In an
18-month period during the early 1970s, 9 million gallons of liquid
industrial wastes including industrial chemicals, sludges, septage
and sludge allegedly were disposed of at the site. These liquid wastes
were either buried in drums or poured directly onto the underlying
soils from an open  spigot on a tank truck.

SAMPLE ACQUISITION
   In order to obtain a sample representative of aquifer conditions,
the monitoring well was evacuated of three volumes of standing
water utilizing a stainless  steel  submersible  pump. Following
evacuation, the sample was collected with a laboratory cleaned and
wrapped dedicated Teflon bottom fill bailer. The decontamina-
tion sequence performed by the laboratory was as follows:

   Non phosphate detergent and tap water wash
   Tap water rinse
   Distilled/Deionized organic free water rinse
   Pesticide grade acetone wash
   Total air dry
   Distilled/Deionized organic free water rinse

   Two sets of  laboratory cleaned bottles were filled in a round
robin fashion allowing zero headspace. A travel blank using organic
free water from the laboratory accompanied sample shipments.
Samples were cooled to 4 ฐ C and immediately transported to the
laboratories under chain of custody procedures. One set of samples
was shipped to a Contract Laboratory Program laboratory for con-
ventional purge and trap analysis and the other set was sent to
Rutgers University for CLSA.

INSTRUMENTAL ANALYSIS
   CLSA  was  accomplished  utilizing a Tekmar CLS-1. The
following experimental conditions  were used with this instrument:

• Purge Time: 1  hr
• Sample Temperature: 40ฐ C
• Trap Temperature: 50ฐ C
• Sample Line Temperature: 60ฐ C
• Trap:  1.5 mg activated carbon
• Desorbtion:  25 pi Carbon Disulfide
               Spectrophotometric Grade (99.9 + %)

GAS CHROMATOGRAPHIC ANALYSIS
   GC analysis was conducted on a Hewlett Packard Model 5840A
under the following conditions:

• Column:  Capillary
            J&W Scientific DB5-30W, Wide Bore, SE-543
            1 fi film thickness
            30m x 0.32 mm I.D.
• Temperature Program:  40ฐ C, 10  min (hold)
                         40-200 ฐC @  4ฐC/min
                         200ฐ C, 10 min (hold)
• Injection:  Spitless
 • Injection Port: 200 degrees C
• Sample Sizes: 2 /il
• Carrier: Hydrogen
• Detector: Flame lonization

  All samples analyzed by the CLSA procedure were spiked with
three internal standards for the purposes of internal calibration.
Some samples were spiked with analyses to establish matrix spike
recoveries.

MASS SPECTROMETRY
  A Carbon Disulfide blank, a 2.0 /tg/1 standard and a monitor
well sample were all closed loop striped and subjected to GC/MS
analysis to attempt to confirm the elution order for the GC Flame
lonization Detector primary analysis. The following conditions
were used:

• Instrument:  Finnagen 5100
• Column: Capillary
           J&W  Scientific DB5-30W, Wide Bore, SE-54
           1 n film thickness,
           30m x 0.32 mm I.D.
• Injection: Splitless
• Program: 40ฐ C, 10 min (hold)
            40-200 ฐC @ 4ฐC/min
            200ฐ C, 10 min (hold)

RESULTS
  The GC data were consolidated into a format suitable for evalua-
tion, outlining retention times (RT) and area counts for each com-
pound detected. Tentative identifications were  made for each com-
pound based upon the elution order determined by the standards
run and GC/MS determined elution order. Later in the project,
GC/MS analysis was performed on  CLS standards and samples
in order  to confirm the order of elution.
  Calibration curves for the standards were constructed for each
target analyte obtained via  the CLSA and GC runs. Calibration
curves for  tetrachloroethane and ethylbenzene are presented in
Figure 1 and 2. These compounds demonstrated good linearity
while other calibration curves had  some anomalous individual
points. Overall, the linearity was good,  especially when the low
concentrations of the standards utilized are taken into considera-
tion (0.4fig/l, 2.Opg/l, 5.OMg/l, and 10.O/ig/l).
   Identifications were further evaluated for each target analyte and
relative retention times (RRT) were calculated  versus  internal
                    349C7B910
                         AMOUNT X (PPB)
                        Figure 1
             Tetrachloroethane Calibration Curve
 86    SAMPLING AND MONITORING

-------
standards. The RT shifts for each run were tabulated. Statistical
analysis was used to calculate the mean (x) and variance sigma (s).
Acceptance windows for each compound were set at x = 3s, the
99% confidence limit. Acceptance windows  were additionally
calculated utilizing an alternate method also in accordance with
current CLP protocols. The second set of acceptance windows was
based on the equation x = 0.3%. For each acceptance method,
the more liberal window was selected for sample run identifications.
  To evaluate the efficiency of the CLSA protocol, sample matrix
spike recoveries were obtained in triplicate and statistical analysis
was  performed.  Information is presented  for  toluene  and
tetrachloroethane in Tables 1 and 2. Precision and accuracy varied
both within each run and from compound to compound. Quan-
tification calculations were performed for these compounds to
determine the amount of analyte recovery.
          1     23456789   10
                        AMOUNT X (PPB)
                          Figure 2
                Ethylbenzene Calibration Curve
                          Table 1
                Matrix Spike Recovery: Toluene
                        2.0 ftg/1 Spike
                                               Run No.




                                               lib

                                               13

                                               15
                                                                      Table 2
                                                        Matrix Spike Recovery Tetrachloroethane
                                                                    2.0 pg/1 Spike
                                            Recovery             Z Recovery
                                             2.90  ppb

                                             1.73  ppb

                                             1.99  ppb

                                             x-2.20 ppb

                                             s-0.61 ppb
                                            145Z

                                             87Z

                                             99Z

                                          K-110Z
                                                                 be attributed to sample contamination as reagent and trip blanks
                                                                 that were also CLSA stripped and analyzed were satisfactory.
                                                                 Monitoring well split samples that were sent to a U.S. EPA IFB
                                                                 CLP Laboratory all came  be  back  as BMDL  at the Contract
                                                                 Required Detection Limits (CRDL).
                                                                                              Table 3
                                                                                 Detection Limits/Sample Concentration
                                                                                             Toluene
                                                         CLSA Analysis

                                                         BMDL: 0.026 ppb

                                                               0.016 ppb

                                                               0.018 ppb
                                                       Purge  &  Trap Analysis



                                                       BMDL

                                                       (CRDL:  5.0 ppb)
                                                                                        Tetrachloroethane
    Run No.




    lib

    13

    15
   Recovery




 0.92 ppb

2.41 ppb

0.69 ppb

X-1.34 ppb

s=0.94 ppb
Z Recovery




  46Z

 120Z

  35Z

X-67Z
          CLSA  Analysis




Concentration:   0.159 ppb

                 0.344 ppb

                 0.160 ppb
 Purge & Trap Analysis



 BMDL

(CRDL: 5.0 ppb)
                                                                                              Unknown A-2
  Monitoring well samples were analyzed in triplicate and quan-
tification calculations were performed to  determine either the
analyte concentration or the method detection limit (MDL) if the
target analytes were BMDL. Detection limits or the concentration
in the sample are presented in Table 3. These analytes could not
                                                           CLSA Protocol

                                              Concentration:   0.188 ppb

                                                               0.245 ppb

                                                               0.032 ppb


                                                                SAMPLING AND MONITORING     87
                                                          Purge & Trap

                                                          BMDL

                                                          (CRDL:   5.0 ppb)

-------
   On two occasions, CLS samples were transported to another
 laboratory for GC/MS analysis. Samples were CLSA stripped onto
 microparticulate carbon then transported to the laboratory where
 they were desorbed and injected into the GC/MS. The first round
 of analysis was not considered successful as only six compounds
 were recovered in the standards run and none in the  sample run.
 Fearing analyte loss in  transportation, additional samples were
 stripped and transported to the laboratory in dry ice. Upon receipt,
 they were immediately placed into a freezer. The desorbtion and
 GC/MS analysis were repeated. This time seven compounds were
 recovered in  the standards run  as well as several analytes in the
 monitoring well sample. While chromotography was  poor on this
 sample run, the compound order of elution was confirmed.  It is
 noteworthy that substantial compound loss on the sample run oc-
 curred in both instances, possibly  due to the inter-laboratory
 transport.
   Quality Assurance and Quality Control (QA/QC) samples for
 the  project consisted of a  trip blank and a reagent (Carbon
 disulfide) blank which were spiked  with the internal standards,
 subject to the CLSA procedures, and analyzed. These runs were
 encouraging as the reagent blank and the trip blank  were free of
 three analytes detected  in the sample at the pg/1 level.


 DISCUSSION
   The Closed Loop Stripping procedure is a valuable analytical
 tool for analyzing monitoring well samples to achieve  substantially
 lower detection limits than can be provided by conventional U.S.
EPA IFB CLP analysis. Monitoring well samples were analyzed
and compounds were detected at the pg/1 level. These samples were
BMDL at the CRDL on CLP split samples. These results confirmed
our hypothesis that the monitoring well sample demonstrated to
be contaminated at pg/I levels would not be detected by conven-
tional chemical analysis. While the CLSA technique is not being
suggested as a  replacement for current CLP protocols, it provides
the scientist with an additional tool in the evaluation of ground-
water quality.
  Problems encountered with the CLSA technique include the level
of skill required, loss of linearity (especially at  concentrations of
10.O/tg/l  and  greater) and analyte  loss during carbon tube
transport. Future  researchers should ensure that  the micropar-
ticulate carbon traps are properly conditioned and maintained. For
this project, traps were rinsed with the Carbon Disulfide solvent
and baked in a muffle furnace for 30 minutes prior to  use. Even
so,  the traps deactivated over the course of the project which might
account for some of the analyte loss experienced.
  Now that the utility of the CLSA method has been demonstrated,
future research should focus upon refining the technique, especially
with regard to analyte recovery. Additionally, other techniques for
achieving lower detection limits warrant exploration. Currently,
sample preparation  techniques  utilizing purge and trap with
cryofocusing and  subsequent  analysis   via  capillary GC/MS
methodology are under investigation. Further research and develop-
ment on alternate analytical protocols is warranted in order to more
accurately characterize groundwater conditions at uncontrolled
hazardous waste sites.
88    SAMPLING AND MONITORING

-------
                               State  Monitoring Well  Regulation:
                                           Need  for  Consistency

                                                      Charles A. Job
                                                    Gilbert Gabanski
                                            Twin City Testing Corporation
                                                  St. Paul,  Minnesota
ABSTRACT
  State well codes vary widely and all are not specific to ground-
water monitoring. Water supply well codes of some states have
requirements  that are contrary to monitoring groundwater. A
review of 12 upper Midwestern and  Rocky Mountain state well
codes highlights these differences. Depending on specific require-
ments among state well codes, monitoring well installation costs
can double. Need for consistency among states will improve the'
quality of work performed, provide for similar levels of cost for
comparable problems and focus on effective resource allocation
in groundwater protection.

INTRODUCTION
  Monitoring well regulations vary  greatly from state to state
affecting groundwater quality, costs of identifying and remedying
problems, and the allocation of people, equipment and materials
on a consistent basis for solving similar problems. The current in-
consistency affects the groundwater protection and improvement
costs which industry must bear. An obstacle to the solution of this
matter is that groundwater is a resource primarily under state
responsibility.1 This situation can be an opportunity for greater
interstate cooperation.
  The environmental and health agencies of the 12 upper Mid-
western and Rocky Mountain states of Colorado, Illinois, Indiana,
Iowa, Michigan, Minnesota, Nebraska, North Dakota, Ohio, South
Dakota, Wisconsin and Wyoming (Fig. 1) were contacted and asked
to provide their most current laws and regulations relating to the
installation and use of groundwater monitoring wells. Of these
12 states, only Minnesota, Nebraska and Wisconsin have regula-
tions or guidelines specific to monitoring wells. The significance
of groundwater in these  12 states is underscored by the fact that
groundwater supplies over 50% of their water needs, other than
for  power production.2
  This paper reviews the administration, construction, materials,
maintenance, testing, reporting, abandonment and disposal use of
the  wells in these states.  It then discusses costs for differing
requirements  and provide conclusions  as to the need for state
monitoring well regulatory consistency.

ADMINISTRATION
  The 12 states  whose regulations were reviewed differ widely in
their  regulations  for  monitoring wells.  Three  of the states,
Minnesota, Wisconsin and Nebraska, have either regulations or
guidelines specifically directed toward monitoring wells while most
other states have either regulations or guidelines directed toward
all wells. The extent to which guidelines are enforceable is a concern
                         Figure 1
       Upper Midwest States Included in Well Code Review

because of the potential lack of consistent application in practice.
In states  having regulations principally for water supply wells,
objectives differ significantly from protecting  public health to
orderly development of groundwater. The purpose of water supply
wells is to provide safe water for human consumption.  The purpose
of a monitoring well is to provide a means to test the quality of
groundwater and not to  supply water for human domestic use.
While regulations for both types of wells must be protective of
water quality, the  construction,  materials and  operation are
different  to meet different purposes.
  Responsible  agencies  range from  a department  of  natural
resources  in Wisconsin to a state engineers office in Wyoming. The
objectives of these offices  also differ as they relate to groundwater:
at one end, managing groundwater programs and at the other,
administering well installation standards. Most states require some
type of licensing, registration or well permitting, although this
requirement is not consistent across all states.  Some states have
no requirement at all for licensing or  registering water well con-
tractors or engineers but require a state well permit.  Other states
require the licensing or registration of drillers but do not require
permits. Several states require the registration of drillers but have
no  minimum  experience or  examination requirements.  The
experience brought  by drillers to the completion of a monitoring
well may affect the quality of that well and ultimately the data
obtained  from it.
  Penalties for not licensing or registering or otherwise comply-
ing with  construction, materials and  maintenance requirements
appear to be relatively minor. However, in today's world the lia-

                   SAMPLING AND MONITORING    89

-------
                           Table 1
                        Administration
                   l**M"lปlซ  ffFซป •! Mill
     luUi I    rr*>U*   DhrUlM •
     UlvUUMU  RUtSU*   Wiur !*••
     6ซrซrซlft|   It*a4*r4<
                            Table 2
                         Conjunction
             tabUf   •with
   tn  ซ ui iti   rniHi  **bliซ iซ*Uk  All
     r.A.  i*rซ. Fin tvfciu itซuป
                                                                                                                       1C      Tซi
     •D Cmtvrr  Prat
     C*ซ* d ป)-!) febll
     t*4 11-11   b.l
   n &.UIIIM
         :tlซB i*n>* ev
                                                                                frM    u- • **-• 01
      a*lซ*llปi   fntHt CM Xt.rtl    )kM|l*rUi|
      fปr w     f*r •ซ•*  biMtrtu
      Iutซimtซ  rปr^T I*n

      fin III •   OT4*rlr    ItBta       All

      *i^l—     .( .1.1. .  0II1C*
bility issue looms large and suits for recovering damages may be
considerable if wells are not properly in-stalled. Because of the great
demand for the installation of monitoring wells to assess ground-
water quality, companies that operate in many states must keep
abreast of requirements in each state or set high  standards for
quality that will meet all states' requirements. These high standards,
while protective of both groundwater  and the long-term interests
of industry, may not  be lowest cost or  cost-competitive in each
case to satisfy short-term  interests.
CONSTRUCTION
  Construction of the wells affects the results of groundwater
monitoring as much as the materials  and subsequent operation.
A review of the 12 states indicates that Iowa, Michigan, North
Dakota,  Ohio, South Dakota and Wyoming have requirements
restricting wells to be set back or out of zones of contamination
or potential contamination. This requirement runs contrary to the
purpose of monitoring wells for groundwater quality. Whether well
code violations are occurring may be a concern.
  Termination of wells above ground has  helped  to protect the
wells from surface water contamination. However, many states
such as Minnesota, Illinois and Michigan, do not allow for flush
grade well installation which results in an installation problem when
working  in  highway rights-of-way, gasoline stations and urban
areas.  Grouting requirements are  specified in most state codes
including the depth of grout. However, some states have not speci-
fied grouting requirements. Grouting  requirements affect poten-
tial contaminant migration to the  water table.
  Several states require protective casing and protective bumper
posts. This requirement further reduces the possibilities for con-
tamination of the well and precludes  damage to the well casing,
thus affecting both groundwater quality and future use of the well.
States leaving considerable latitude  or providing minimal guidance
on construction can potentially compromise the integrity of wells
for groundwater monitoring. Guidance should be specific so that
wells can continue to be put to their intended use, minimize rein-
vestment in wells and allow the governmental agency involved to
obtain the results it needs  to direct its programs.

MATERIALS
  A wide range of requirements exists for well materials. Most
states permit polyvinyl chloride (PVC) or thermoplastic casing, steel
casing and, in some cases,  Teflon casing. Even within the casing
types, there  are  differences:  Minnesota will not permit flush-
threaded  PVC joints; Nebraska recommends  avoiding solvent-
based cement for PVC casing; and Wisconsin  guidelines do not
allow solvent-based cement for joints.
                            Tabk 3
                         Well Materials
90    SAMPLING AND MONITORING

-------
  Most states have requirements for weights and dimensions of
casing. This requirement is important for the long-term use of the
well. Locking cap requirements, while appearing to be insignifi-
cant, play an important role in reducing vandalism and liability
in monitoring well contamination. Most states have grout mixture
specifications. Those states  which do  not have  grout mixture
requirements may be  affecting the long-term use of those wells for
consistent groundwater quality monitoring.
  Discharge control  from the well is important for two reasons.
The first reason is to  preclude backflow into the well of contami-
nated water if the well is routinely used for groundwater quality
purposes. Secondly, protection from freezing is important in the
overall  system integrity and long-term use as frozen  lines can
contamination are penetrated and not sealed. Six states specifically
require reporting of well abandonment. Procedures for test hole
and well abandonment affect for water quality tracking and needs
for redeveloping sites.
  Five states preclude wells from use for waste disposal. Two states,
Ohio and Michigan, will allow waste  disposal but only  on the
approval of the agency. Waste disposal obviously can compromise
the use of wells for groundwater quality monitoring.
                             Table 5
                  Abandonment and Disposal Use
damage pumping systems.
MAINTENANCE, TESTING AND REPORTING
While many items could be cited for maintenance, testing and
reporting, several have a direct bearing on the use of wells. Disin-
fection is required for water supply. However, disinfection with
chlorine is not consistent with monitoring well use. Yet, most states
indicate that their water supply well regulations are applicable to
monitoring wells. The intended use of the well is inconsistent with
applicable well codes. The operators of the wells may be violating
the law in these states.
Table 4
Maintenance, Testing and Reporting
Correction Yield Veil Veil
Stste Disinfection of DAUB* Text Ssnoles Records

Colorado Yes 	 Yes 	 	
Iowa Yes 	 	 	 	
Hichie.sn Yeซ 	 	 	 Yes

Minnesota Alternative Yes/72 hours Ho Yes/30 dajrs Yes/30 dsys
Methods
Nebrsska Yes
Ohio Yes 	 	 	 Yes/30 
-------
COST COMPARISON
  In examining the requirements of different states, it is clear that
monitoring wells in one state may be very different from those in
another state. One method of quantification is to assess the costs
of installing different types of monitoring wells. Table 6 describes
two wells and associated costs. A 50-ft stainless steel-cased well
with maximum surface protection and grouting is approximately
twice the cost ($1,950 to $2,100) of a PVC cased well with grout
near the ground surface and no surficial  protection ($1,010  to
$1,160). Using the information from Table 6, one can extrapolate
costs for a PVC cased well with maximum protection to be about
25 % higher. This is a significant  additional cost, should a large
number of wells be required. This added cost includes additional
time, equipment, and materials.
  The costs for disinfection, sampling, aquifer performance testing
(such as yield tests) and reporting could increase the difference in
costs, even though some of these  activities may not be essential
for monitoring wells. If they are required  by law or regulation,
however,  they are part of the total cost of well installation and
performance and should be evaluated.

CONCLUSIONS
This review of state well codes indicates that:

•  State well codes vary widely
•  Experience and training requirements for water well construc-
  tion do not exist in many states
• Portions of some well codes  relating  to  construction, well
  materials, maintenance and abandonment may not adequately
  serve the purpose of groundwater monitoring wells
• Specific requirements on location, termination, grouting, pro-
  tection, disinfection, damage and testing should be reconsidered
  with respect to the different purposes of monitoring versus water
  supply wells
• Depending on requirements, costs can range by a factor of two.

  More  consistent monitoring  well  codes among  states can
eliminate confusion in industry thereby improving the quality of
work and of the resource. This may be particularly important when
interstate or federal concerns are raised later in a project such as
a location subsequently becoming a federal Superfund site. More
uniform monitoring well codes can also provide for similar levels
of cost for responding to groundwater problems. This consistency
will enhance  the effective allocation of people, equipment and
materials for this important aspect of groundwater protection and
improvement.

REFERENCES
1. US. EPA, Office of Ground-Water Protection, "Ground-Water Pro-
  tection Strategy," Washington. DC, August, 1984.
2. U.S. Geol. Survey, "National Water Summary," Water Supply Paper
  2275. Washington, DC, 1985.
92    SAMPLING AND MONITORING

-------
                      Quality  Assurance  for  the  Field Laboratory
                                                 B. Chris Weathington
                                           Hittman  Ebasco Associates  Inc.
                                                 Columbia, Maryland
ABSTRACT
  The use of field laboratories during remedial investigations, site
cleanups and facility decommissionings has become an accepted
practice. The importance  of planning  and liaison between the
laboratory providing the services and the engineering firm using
the field laboratory is important. Pre-mobilization considerations
include  analytical procedure selection,  personnel requirements,
number of analyses per day, reporting criteria, field laboratory/site
management liaison, instrument needs and quality control/quality
assurance (QA/QC).
  Upon completion of the mobilization phase, the field labora-
tory  and site managers establish the daily operational protocols
for sample submission, chain of custody and reporting. During
the planning phase, decisions are made with respect to the quality
of data the  field laboratory will  provide. Data verification and
validation requirements are  discussed. Results from the field labora-
tory  may vary with respect to the level of quality assurance used.
  Three QA/QC levels are described which should be applied based
upon a client's engineering or litigation needs. Level I has criteria
for simple and rapid site screens to determine  the extent and type
of contamination with a qualitative estimate of concentration levels.
Level II includes more precision and accuracy controls for better
estimates of concentration.  Level III provides data with the highest
criteria for accuracy and precision. Level III QC procedures would
include  complete chain of  custody, sample security, calibration,
method blank criteria, data certification, confirmation of identi-
fication, complete documentation and data validation. The posi-
tive and negative factors for using these various  levels are discussed.

INTRODUCTION
  Demand for field laboratory operations is increasing as remedial
investigation, emergency response and cleanup teams become aware
of the services a field laboratory can provide. These services include
analysis for volatile and semi-volatile organic contaminants, metals
and inorganic indicator parameters. Sample matrices include water,
soils  and multi-phase hazardous wastes. For the field investigation
team, the most important service the laboratory can provide is rapid
turnaround. Rapid turnaround, once experienced by a field inves-
tigation or cleanup team is habit-forming.
  What makes  the field laboratory so addictive is its immediate
responsiveness to a site manager's needs. He has immediate answers
concerning where to sample next, where the hot spots are and when
to stop  digging or collecting samples.  During remedial investi-
gations, he can, with on-site analysis, quickly mark areas at the
site as clean and then go on to the suspect places. During cleanup
activities, he can quickly screen samples for removal or treatability.
  The field  laboratory providing  these rapid services is under
tremendous  pressure to assure not only a rapid turnaround of
results but a result that is scientifically correct. To do this, the field
team and field laboratory must develop the data quality require-
ments through a project quality assurance and quality control plan.
The various aspects and requirements for quality assurance in the
field laboratory are described in the sections that follow.

PRE-SITE PLANNING
  For the home laboratory providing mobile analytical services to
the field engineering firm, a number of questions must be answered
to ensure the appropriate staff, instruments and level of quality
assurance  are in place when sampling starts. The first and most
important question is: "What is the objective of the investigation?"
This is the question most frequently forgotten when developing
the project analytical methods and QA/QC plan. To give the best
analytical  product, both economically and rapidly,  the  project
objectives  should be clearly defined. Objectives may include:

• Site investigation
• Site cleanup
• Enforcement
• Feasibility study

  With these objectives and a target list of compounds in hand,
the laboratory can begin its evaluation of the analytical methods
and quality control  features that best fit the investigation.

ANALYTICAL METHODS
  Choosing the correct analytical method for a site is dependent
upon whether the data are for qualitative, semi-quantitative, quanti-
tative or litigation/enforcement quality information. Each type of
data has advantages and disadvantages in a mobile laboratory sit-
uation. Sources for these methods include the U.S. EPA's manuals
SW-846 or 600 series methods, publications of the ASTM, AOAC,
APHA or  the scientific literature. In some situations, a vendor's
procedures or laboratory developed methods may be most ap-
plicable.
  Some procedures include standard quality control  practices as
part of the  method documentation and may be applied in the mobile
laboratory. However, it is frequently necessary to modify the QC
procedures in a fashion to increase through put without jeopar-
dizing the analytical result. For example, qualitative analysis devised
from an U.S. EPA Gas Chromatography method may require, in
the original document, analysis of calibration standards for quan-
titation and  identification as well  as duplicates,  spikes,  blanks,
check samples and second column confirmation. But for a qualita-
                                                                                     SAMPLING AND MONITORING     93

-------
live result, the QC requirement would be a blank, a calibration
standard for identification and establishing the target reporting level
and 5% of the samples confirmed for ID. These QC procedures
would be performed daily except for the 5% confirmation which
may be performed by the mobile laboratory or at a fixed labora-
tory on a weekly basis.
  Modifications made in a procedure  should be validated by the
laboratory before use in the field. The field laboratory should not
have to make modifications in procedures while in the field. Matrix
problems  should be handled by the  home laboratory.  In any
situation where the field  chemist has to modify a procedure, he
must  fully document what is done and how the  method was
validated.
  Semi-quantitative, quantitative and litigation-quality analyses
usually are methods that have been  fully validated by the agency
or organization that documented the method.  These procedures
should not be modified without backup data from the laboratory
to ensure the results are correct. Modifications taken in  regula-
tory procedures may invalidate the results and, therefore, specific
approvals should be obtained before  proceeding.

DOCUMENTATION
  To  ensure the most efficient use of a field laboratory, documen-
tation packages should be prepared prior to leaving for the site.
The documentation packages should include forms for:

  Sample log-in
  Sample preparation
  Preparing  calibration table
  Recording blank  analyses
  Summarizing QC results
  Reporting  sample results
  Documenting chain of custody

  All of these forms are not required  for every site,  but are
dependent upon a  specific site's needs.  Using prepared forms
reduces documentation efforts by the field laboratory staff and
thus increases sample throughput. The forms should be clear and
simple in the display of the information needed to document and
track  results. Figure 1 is an example of a prepared field logbook
page.
                        fltlO KBOMTORr
                           PESTICIDE
                         PRtPtMllOX IOC
             DATE:

           AKALTST:

   EITRACTINC SOLVEHT:

  SOLVENT EICHAKGCD TO:
   PACE Nl"1[ซ:

NOT CROOK NUMBER:

   REVIEWED BT:
TOTAL VOL.
HEAI KEICHT tllB.
SAMPLE ID 1 (v>) SOLVENT
FINAL
VOL. (A) 1MI. VOL. FINAL
(OS ANAL. (it) VOIUMC I*,)
1
2
3
4
5
6
                           Figure 1
                  Example of a Field Logbook
  Since most field laboratories have from one to three persons
performing the analysis, documentation of results may not be as
complete as needed due to the pressures to report analyses. Using
formatted logbooks does help, but it does not result in a perfect
                              documentation package. When possible, the field laboratory should
                              have one person assigned to check all paperwork for completeness.
                              It is very difficult to backtrack after leaving the site.
                                Chain-of-custody documentation can be very difficult to ensure
                              if sample splitting is required as a part of the QA program. This
                              is particularly a problem due to the lack of space within the field
                              laboratory for storage  of samples. Outside  storage is acceptable,
                              but the samples must be maintained in a locked ice chest or refri-
                              gerator. Outside storage cannot be a guaranteed means of assuring
                              chain of custody  since access is possible when no one is on-site.
                              Furthermore, cooperation between sampling technicians and the
                              field laboratory is a must, otherwise sign-offs may not occur.
                                If at all possible, documentation requirements should be kept
                              to a minimum. The most favorable situation is having only three
                              documents to fill out.  These documents  would include:

                              • The chain-of-custody form—this form would have the sample
                                ID, analysis required and space for result of analysis
                              • The sample preparation logbook—this notebook  would  have
                                space for all  preparation steps
                              • Multi-analyte report form—this form would have a list of GC
                                target compounds (i.e., pesticides,  base neutrals) and space for
                                results.

                                Instrumental output would be stored in  a folder by date with
                              the  original forms. Data reported to the site investigation team
                              should be photocopied, if possible. As an alternative, all original
                              data may be turned over to the site  manager.
                                When at all possible,  the field laboratory should have a computer
                              available  for reporting and data  reduction  purposes. Besides
                              providing a better looking report, the computer can perform the
                              calculation of results and store data in a form that can be used
                              by the investigation team for report preparation.
                                Documentation of results for litigation/regulatory purposes is
                              the  most time-consuming and rigorous situation for the field labora-
                              tory. The quality  control documents, as well as routine reporting
                              requirements must be clearly traceable throughout the analytical
                              process. The results of each step in analysis must be logged, dated
                              and initialed. A second person should check results and initial and
                              date that the checking function took place. Deviations in documen-
                              tation or analytical procedures must be noted with valid scientific
                              reasons and data to back up any changes. It  is recommended, that
                              for  economy and  efficiency in field laboratories, litigation quality
                              data not be expected or attempted regardless of the need for timely
                              results.
                                                         Table 1
                                    SUMMARY OF QUALITY CONTROL PROCEDURES
                                                                    Type of Analyses

                                                                    Qualitative
                                                                    Semi-quantitative
                                                                                         Mini.
                                                                                                Controls
                               Quantitative
1 Blank

2 Calibration standards



1 Blank

2 Calibration standards

SI Duplicate/samples

SI Spike/samples



1 Blank

Initial standard curve (beginning of the program)

2 Calibration standards

SS Duplicate/samples

51 Spike/samples

1 Performance evaluation  sample
94     SAMPLING AND MONITORING

-------
QUALITY CONTROL
  The quality control procedures the mobile laboratory uses should
be based on:

• The type of method (i.e., GC, AA, UV-VIS)
• Ultimate data usage
• Sample throughput (efficiency)

  The minimum control any method should have is a blank analysis
to ensure there is no laboratory contamination, a calibration/check
sample to establish that each instrument is operating correctly at
the beginning of the day and another calibration/check sample at
the end of the days runs to ensure continued correct operation.
Table 1 lists the various types of control samples that  might be
required for several examples of analyses.
Type of Analysis

Qualitative


Semi-quantitative




Quantitative
Minimum Controls

One Blank
Two Calibration standards

One Blank
Two Calibration standards
5% Duplicate/samples
5% Spike/samples

One Blank
Initial standard curve
    (beginning of the program)
Two Calibration standards
5% Duplicate/samples
5% Spike/samples
One Performance evaluation sample
  Analysis of control samples enables real-time data evaluation
to assure the  correct operation  of instrumentation or analysis.
Criteria  for control may be based upon the laboratories' own
operational control limits or those established by U.S. EPA. In
some cases an arbitrary go no-go limit for percent recovery or
percent difference may be acceptable. For example, ฑ25% for a
difference between duplicates or ฑ 20% for a % recovery might
meet the requirements of the field investigation team that require
semi-quantitative or quantitative data of a specified quality. Some
type of corrective action by the field laboratory would be neces-
sary for any analytes which do not meet these criteria. However,
since the above criteria are arbitrarily chosen the analyst has some
leeway in deciding whether an analysis is not working in accordance
with established standards.
  During mobile laboratory operations, quality control is best used
by the analyst on a real-time basis  to make immediate correction
of the analytical system. There is no time for long-term decision-
making with regard to the next step in fixing a problem. Nor is
there a  redundancy of instrumentation  or parts to correct a
problem. The analyst must evaluate a control samples instrument
response values with any previous data before  continuing with
further analysis of samples. The duplicate and spike control sample
results must be evaluated for matrix  problems. Should matrix
problems be indicated by control sample results, the field labora-
tory usually  has alternate  procedures available to  correct the
problem. Otherwise, the field laboratory may submit those samples
which might cause matrix problems to the home laboratory for
analysis.
  In many cases time constraints result in the quality control being
performed but not acted upon. The analyst doing  the work may
have to rely upon past experience in  accomplishing an analysis while
documenting the control steps. The field laboratory may only be
scheduled for 2 weeks at a site and therefore, the results of analyses
may not be fully validated on-site.
VALIDATION AND VERIFICATION
  Field laboratories are limited in the number of personnel and
space available to fully verify and validate results. The site manager
and  field chemists must,  in  most cases,  check the data  for
completeness. Data packages prepared by the field chemists may
be nothing more than instrument printouts, logbook notes con-
taining calculations and QC results. As such, these data packages
have the minimum documentation for a reviewer  to check  the
results.
  Because of the short period of time on many sites, data review
is in the  hands of the  analyst. Some sites provide  the time and
money for an outside audit of the data being produced by the field
laboratory. The outside auditor may be someone from the client
requiring the work, the U.S. EPA  Regional QA Office, the  site
investigation team's QA officer or  the laboratory QA personnel
providing the field services.
  Outside auditors cannot provide full validation of  the field
laboratory,  but they can provide supporting evidence that pro-
cedures are being followed and that the results "appear" to meet
the data  quality objectives  of the project. Therefore, the valida-
tion process for  data which  is usually  performed  by a fixed
laboratory should never be expected in a field laboratory situation.

PERSONNEL
  The most important element in the field laboratory is  the
personnel chosen to perform the analyses. The key individual is
the lead chemist, whose responsibilities include liaison with the field
team, sample preparation,  analysis, "house-keeping," reporting
of  results  and quality assurance/quality  control. Classically,
QA/QC  is not and should not  be performed by an individual
responsible for actually performing the analysis.  Economics and
laboratory size, however, rules out an individual with sole responsi-
bility for QA/QC.
  Personnel in the field laboratory should be highly qualified and
experienced with the analytical methods and matrices that are to
be analyzed. The lead chemist  may have an organic or inorganic
analytical background depending upon the site. A record of the
chemist's background and expertise  should be available to indicate
the qualification of the lead  chemist  and technicians. Since the lead
chemist must be capable of handling analytical or instrumentation
problems by taking effective corrective actions, his credentials
should be as broad as possible. Some field situations may require
the lead  chemist and technicians be trained prior to entering  the
field by performing check sample analysis. Any training that takes
place should be documented in the QC files.
  At the technician level, the personnel may be entry level or have
no specialized experience. The tight  conditions of the field labora-
tory and repetitive nature of many procedures permits close super-
vision by the lead chemist. Technician level personnel perform
many  of the  less demanding  analytical  procedures of sample
preparation, documentation,  sample  control and inventory.
Documentation of a technician's experience and capabilities should
be available and reviewed before field work begins.

DATA QUALITY LEVELS
  Data quality levels are a  means of describing the QA/QC that
will be provided for a site  investigation.  Three levels of quality
should be available for the field laboratory. Each  level is based
upon the data quality objectives: Level  1  provides screening;
Level 2 provides semi-quantitative estimates; and Level 3 provides
quantitative results.
  Most  field laboratories  should fall into Level 1 or Level 2.
Level 1 screening data are best  suited for emergency response  and
cleanup  situations. The results from Level  1 data are qualitative
in nature. The results of a particular analyte is reported as a greater
or less than value.  The detection limit for each analyte assayed for
is determined before entering the field. The cutoff value should
be established daily by a calibration standard at twice the target
detection limit, if possible. A  site investigation team using these

                     SAMPLING AND MONITORING     95

-------
data can quickly determine the clean versus hot spots at a site.
Composite samples can be prepared at the hot and clean spots and
sent to a referee laboratory for confirmation. Data produced at
Level 1 can also be used at sites where well installation is neces-
sary by determining the most likely areas of contamination. Level
1 data should not be used to write  definitive reports.
  Level 2 objectives for data include an estimate of the  concen-
tration of target analytes at the site. Accuracy and precision  of
the data are calculated by the field laboratory and documentation
packages are prepared that can be validated at a later date. Sample
control through chain-of-custody can be performed by the field
laboratory team. However, sample capacity and turnaround may
be reduced by the field laboratory. Decisions for cleanup or further
investigation can be made from these data. At least 10% of the
samples should be sent to a referee laboratory for confirmation.
  Level 3 provides data of quantitative or regulatory nature. The
results are fully tested for accuracy and precision. Chain-of-custody
and sample control  are maintained, results are validated to the
extent possible, analyses are confirmed on-site and the reports from
the laboratory should meet close review for quality and validity.
Again, the productivity of the laboratory is significantly reduced,
compared to other levels. Confirmation of 10% of the samples
at a referee  laboratory should be performed. Producing data at
Level 3 in the field laboratory should only be considered in those
situations where turnaround of a final report in the shortest time
possible is required. The Level 3  field laboratory should be given
adequate time on-site to complete and report all analyses before
leaving the site.
CONCLUSION
  Quality control in the field laboratory should be formulated on
the objectives of a site investigation. Screening for target analytes
is the most appropriate use for the field laboratory and can provide
the most efficient and economical results. In some situations, the
field laboratory can be equipped with the personnel and instrumen-
tation to provide results that are on a par with a fixed laboratory,
but both efficiency and economy may  be lost.
96    SAMPLING AND MONITORING

-------
            Effects  of Environmental  Variables  on  Soil Gas  Surveys
                                               Louis S.  Karably, PE, PG
                                                    Kevin B. Babcock
                                          Law Environmental Incorporated
                                             Government Services  Division
                                                    Atlanta, Georgia
ABSTRACT
  Soil  gas sampling  is an effective method for detecting and
defining the location and extent of VOC contamination in the sub-
surface. This site assessment technique has been used to select the
optimum locations for monitoring wells and  site remediation
approaches at numerous CERCLA sites. During the course of these
site assessments, investigators have noted significant variations in
soil gas sample point readings over time. Environmental variables
are believed to be the primarily cause of these variations.
  This paper presents our recent at a VOC-contaminated site,
where soil gas sampling was used to define the extent of contami-
nation at a former fire training area. This example site illustrates
the environmental variables which complicate the collection and
analysis of the soil gas data. Forty permanent soil gas monitoring
stations were installed around the suspected source of contamina-
tion at the site. Variations in soil gas sample station readings over
time were observed. Subsequent analysis showed that although the
VOC concentration levels for each respective sampling date were
different the general  trend of the contaminant  levels and plume
geometry were roughly the same. These results suggest that levels
of measurable VOCs fluctuated due to both temporal and climatic
changes. Another test at this site addressed the variation in VOC
readings as a function of extraction time.
  These trends parallel the expected behavior of VOCs based on
then- chemical and physical properties. VOCs tend to be present
in soil  gases above the source of contamination because of their
inherently low water solubilities and specific gravity and high vapor
pressures. In this case study, soil gas concentrations were affected
by the varying climatic conditions that control gas-phase diffusion:
temperature  and  infiltrating groundwater. Because  changing
climatic conditions dominate shallow soil gas phenomena, the soil
gas concentrations and the gas concentration profile varied with
time.
  Physical properties of VOCs and field observations imply that
readings taken over a period of time or during changing climatic
conditions should be calibrated back to permanent soil gas stations.
In light of changing climatic and temporal conditions, multiple soil
gas readings at permanent soil gas stations should be considered
in order to establish  meaningful trends for  site assessments.

INTRODUCTION
  Soil gas sampling and analysis is an evolving tool for the evalu-
ation of hazardous waste sites.  This investigation technique has
effectively detected and defined the location and extent of VOC
contamination in the subsurface. Soil gas investigations have been
used to estimate the extent of VOC contamination and to select
the optimum locations for groundwater monitoring wells and site
remediation approaches on CERCLA sites.
  Soil gas sampling and analysis is not yet conducted under univer-
sally accepted and regulated protocols. Each investigator is cur-
rently developing and using his own  sampling procedures and
analytical techniques. No consistent or standard sampling metho-
dology addresses the affects of site environmental conditions or
the impact of sampling technique on measurable concentrations
in the soil gas. As a consequence,  the resulting analysis does not
meet the quality assurance standards necessary to compare sites
to criteria like those established for groundwater. Unlike U.S. EPA
and state groundwater criteria, soil gas is not directly regulated
at this time.
  Although soil gas sampling has been effective in many geologic
settings, it has been highly dependent on environmental factors.
The precision of the technique  is controlled by local climatic and
hydrogeologic conditions as well as the sampling technique. A set
of experiments was conducted at a fire training area which was
contaminated with hydrocarbons and solvents. It was used at this
site to determine whether the soil gas readings represented a soil
or aquifer contaminant source. The soil gas analyses were con-
ducted on-site from probes which were augured into the soil. The
resulting soil gas data, recorded on a photoionization detector
(PID), reflected a soil contaminant source.
  Other experiments at this site addressed the fluctuations of the
total organics in the soil gas as a function of time, temperature
and climatic conditions. Preliminary results suggested that the soil
gas concentration increased during the early  afternoon  and
decreased in the late afternoon roughly correlating with the changes
in the temperature during the  day. Environmental variables are
believed to be  the primarily cause of these variations.

Sampling Techniques
  Soil gas sampling has been performed by a variety of techniques,
including driven perforated probes, driven hollow probes, Surface
Static Trapping (SST), augured permanent stations and other
hybrid techniques. It generally is recognized that soil gas sampling
was developed by the oil and gas exploration community to find
hydrocarbon deposits1-2. The SST/Petrex sampling technique was
derived for this application and involves the use of an activated
carbon trap which accumulates hydrocarbons over a period of time
and is then collected for laboratory  processing  and analysis3.
More efficient techniques have been developed for hazardous waste
site investigations and underground storage tank testing4 More
recently investigators have used  augers to introduce soil gas
sampling probes5. During the same time-frame, other investi-
                                                                                               FIELD SCREENING    97

-------
gators were developing driven perforated probes and hollow probe
techniques6'7.
  Each technique had its own strengths and weaknesses. The driven
perforated and hollow probes offer the advantage of fast installa-
tion time; however, repeated analysis of same probe locations over
an extended period of time is not possible. The SST technique
requires that the sensor be left in the ground over a period of time,
and thus requires a relatively long response time. Augured perma-
nent stations offer repeatablc locations over an extended period
of time; however, the installation of the probes is relatively  slow.
While some methods provide relatively expedient results perma-
nent probes offer the capability of repeated sampling and use as
a long-term  monitoring network. Due  to the slow  process of
molecular diffusion, long-term changes in soil gas concentrations
will be readily evident using this technique.

Analysis Techniques
  Once the soil gas is extracted by one of these sampling techniques,
it is analyzed in the field or transferred to a laboratory for analysis.
The analysis method ranges from the simple hand-held photoioni-
zation detectors (PID) and organic vapor  analyzers (OVA) to
sophisticated field gas chromatographs (GC) and laboratory  grade
instruments, such as the GC/MS.
  Soil gas samples can be analyzed  with three different levels of
rapidity  with corresponding increasing accuracy of results. For
immediate results, a portable photoionization detector (PID) or
organic vapor analyzer (OVA) can be used to sample total  VOC
levels. Within a few minutes, a rough estimate of total organics
present in the sample can be  made. For a more detailed assess-
ment, GC or GC/MS analyses can provide relatively accurate iden-
tification and quantification of contaminants, especially when there
is prior knowledge of expected contaminants, and can be completed
in less than 1 hr.  As the level of analysis becomes more sophis-
ticated there is an increase in the cost of supplies and personnel
to operate the equipment. Most investigators during the  early
development of the soil gas  technique noted the influence of
sampling depth, soil moisture and permeability and the effect of
constituent-specific effects*.
                                      PHOTOIONIZATION
                                          DETECTOR•
       IM" POLYETHYLENE  (HOPE)
           SAMPLING TUBING
                  V
                                PERISTALTIC
                                   PUMP
                              FILTER
CASE STUDY
  Forty permanent soil gas monitoring stations were installed
around the suspected VOC-contaminated zone for an USAF fire
training area. The soil gas sampb'ng technique at this site consisted
of hand augering shallow borings 4 to 6 ft into the vadose zone.
High-density 1/4 in. diameter polyethylene tubing, with a filter
device on the end, was installed and backfilled. The above ground
end of the  probe was fitted with a cap and allowed to stand for
a minimum of 24 hrs to permit normalization of the VOCs in the
soil gas station.
  Soil gas was extracted via a peristaltic sampling pump (Fig 1).
The soil gas was evacuated from the soil at a rate of 1.51/min into
a 0.5 I Tedlar sampling bag. The probe of the PID was connected
to the sampling  bag and the level of VOC vapor was recorded.
A total of 20 to 30 min was required at each station to auger the
hole,  install the  probe and collect and analyze the sample.
  The quality control used in the analytical procedures consisted
of recalibrating the PID with the calibration gas  every five sam-
ples. Soil gas samples were qualitatively analyzed on-site using a
Photovac TIP TM PID. The TIP PID used in the survey mode
served as an indicator of total ionizables present in the vapor phase.
Since compound separation and identification was not possible with
this instrument, total concentration was measured as an isobuty-
lene response equivalent ppm. An up, gradient location from the
suspected contamination was selected as a background sample point
for the contaminated area.  At this control boring,  only trace
amounts (1-2 ppm) of total VOCs were detected in the soil gas.
  The 40  soil gas monitoring stations were used to  detect and
measure volatile organic gas levels in the soil. The soil gas survey
was able to delineate several sources of contamination in the dis-
tressed region of the site. Real-time VOC versus  depth profiling
was performed to determine whether soil gas values were represen-
tative of soil or groundwater contamination. The distribution of
contaminants increased in concentration to a depth of approxi-
mately 3 to 4 ft, which then decreased, indicating a shallow sub-
surface contaminant source.  The depth to the water table was
approximately 11 ft and probes were augured to  a depth of 4 to
6 ft. The soil was a relatively homogeneous clayey  sand over a thin
clay layer at 5 to 7 ft depth.
  A soil gas plume was mapped for several days  under different
temporal and environmental conditions. The results of analyses
from these  stations indicate an elongate area with localized centers
                                                                                                                            o
                            Figure 1.
                   Soil Gas Monitoring Station
                            Figure 2.
          Soil Gas Isopleths Following a Cool, Rainy Period
 98    FIELD SCREENING

-------
of contamination in the unsaturated soils in the vicinity of the fire
ring. Data collected over a period of time and under differing cli-
matic conditions (T = 60 ฐ F, T = 90 ฐ F) showed that although the
VOC levels for each respective sampling date were different the
general trends of the contaminant levels and plume geometry were
similar over time (Fig 2, and 3). The distribution of total organics
found in the soil gas appears to  be relatively localized.
  Groundwater samples collected  from the monitoring wells in the
vicinity of the fire ring show that  the contaminants are migrating
through the groundwater system. The monitoring well located near
the center of the site (MW 112) contains the highest concentration
of constituents in the groundwater. Concentrations decrease toward
the southwest (up-gradient). Values decrease more gradually toward
the east and north, indicating possible down-gradient migration
and dispersion of the contaminants from the fire ring area. Based
on the soil gas  survey  and the groundwater flow direction, it is
believed that the contamination does not extend far from the cen-
tral monitoring well toward the west and south (Fig 4).
                                                       o
                                                       SCALE IN FEET
                           Figure 3
           Soil Gas Isopleths Following a Hot, Dry Period
The distribution of contamination appears to be affected more by
the shallow stratigraphy of the site than the groundwater flow. This
distribution may be attributed to the two layered geological system
with the less permeable clayey layer attenuating the source con-
tamination to the groundwater.  Thus,  contaminant migration
appears to be controlled by infiltration down to the first clay layer
where the constituents then move westward and slowly leak through
the clay layer into the shallow unconfined aquifer.
  The soil samples obtained and tested, which were located in an
area of relatively high soil gas readings, show higher levels of
petroleum  hydrocarbons which  relate  to  the  soil gas survey
numbers. The high readings from the soil  gas survey may represent
pockets of petroleum hydrocarbons, such as tars and semi, vola-
tile long chain carbon molecules, rather  than the volatile organic
contaminants found in the groundwater. The contaminant materials
at the site appear to be concentrated in  the upper soils and near
the top of the shallow aquifer. Soil samples from the wells show
a rapid decrease in petroleum hydrocarbons with depth.

DISCUSSION
  Soil gas samples showed a deviation in  concentration levels with
samples collected in the morning and in the late afternoon. Single
sample point VOC readings ranged from 1 to 15 ppm. These vari-
ations in concentration represent changes in soil  gas probably
attributable to vapor  releases in the area with highest values
measured in the late afternoon. Volatilization  was generally
enhanced by dry conditions. Actual quantities of volatiles removed
by each pathway are strongly affected by the climate and soil type
at the site.
  Aside from all the variabilities that can arise due to the presence
of complex soil conditions, sampling methodology alone can create
significant variations in the data. An experiment which addressed
the variation in VOC  concentrations as  a function of extraction
time was performed.  Repeated sampling at individual soil  gas
stations showed a marked increase in a total organics concentra-
tion levels as more samples were taken. This time-based relation-
ship  was for the most part  curvilinear until it normalized at  the
steady-state concentration of soil gas (Fig 5). This finding suggests
that an optimal volume of soil gas needs to be evacuated to obtain
a representative sample. However,  this relationship may vary from
location to location and is  site-specific.
a.
a
Z
O
tr
K
0
0



o
cy —

o
1"
o


8 r
^""~~ —
--^^
/^
/ป
STEADY
n



KEY
STATE
n
i



O 11-11-86 DATA
1 • 12-5-86 DATA

I I I
0.0 2.5 5.0 7.5

\
10.0

i
12.0 15
                            Figure 4
           Groundwater Constituents Concentrations (PPB)
                      PURGE TIME (minutes)

                          Figure 5
             Effect of Purge Time on Concentration


  Data validity was evaluated by assessing analytical and field vari-
ables utilizing  long-term monitoring  stations,  which  allowed
repeated sampling over the course of the study period. The tech-
nique was compared by correlating to several stations. Precision
analyses performed on the sampling episodes indicate low analyti-
cal variables and relatively high field variables. Field variability
was evaluated by examining relative differences in concentration
                                                                                                    FIELD SCREENING    99

-------
for all samples at the same probe. Field variables can be divided
into  two components: environmental and sampling.  Although
sampling variability can not be quantified as separate from environ-
mental and temporal variables it can be minimized by adhering
to effective and consistent field techniques.
  The reproducibility of the  method  was checked.  Repeated
sampling was performed at the site on successive weeks. The second
sampling episode was performed after a rainy period which lowered
the soil gas concentrations observed at the stations. Selected borings
were sampled within  the same day to check for reproducibility.
Tests showed variability between samples throughout the day as
well  as on different days of sampling. It is apparent that the varia-
bility at this  site is attributable to the environmental conditions
which far outweigh analytical considerations.
  These trends parallel the expected behavior of VOCs based on
their chemical and physical  properties.  VOCs  tend to be present
in soil gases above the source of contamination because of their
inherently low water solubilities and specific gravity and high vapor
pressures. Their tendency to escape from the groundwater into the
soil gas is a function of their  concentration in the groundwater and
soil,  their  aqueous solubility, specific gravity and  their vapor
pressure. The VOC's volatilize out of the groundwater and soil
into the above soil gas and move upward by molecular diffusion.
VOC soil gas concentrations were affected by the varying climatic
conditions that control gas-phase diffusion namely temperature
and  infiltrating groundwater. Because changing climatic conditions
dominate shallow soil gas phenomena, the soil gas concentration
profile, as well as  distribution profiles between the three phases,
will  vary with time.  Volatilization was  observed to be enhanced
by dry, warmer conditions.

CONCLUSIONS
  Physical properties of VOCs and field observations imply that
soil  gas readings taken over a period of time or during changing
climatic conditions should be calibrated back to  permanent soil
gas stations. In light of changing environmental and temporal con-
ditions, multiple soil gas readings at permanent soil gas stations
should be considered in order to establish meaningful trends for
subsequent site assessments.
  Sampling at soil gas stations should be based on site-specific soil
gas extraction protocols. Soil gas readings are a function of actual
VOC concentrations, sampling depths, environmental conditions
and the soil gas purging history. Ignoring any of these  variables
can lead to an eroneous analysis of the soil gas distribution at a site.
  Soil gas surveys conducted in phases are subject to substantial
baseline shifts due to seasonal effects. These can be primary effects,
such as those related to temperatures, or secondary effects, such
as those related to precipitation and subsequent infiltration effects
on soil gas concentrations.

REFERENCES
1.  Horvitz, L., "GeochemicaJ Exploration For Petroleum," Science. 229,
   1985. 821-827.
2.  Hickcy, J.,  "Preliminary Investigation of an Inlegrative Soil Gas Tech-
   nique for  Petroleum Exploration," Master  of Science Thesis from
   Colorado School of Mines, Golden, CO, 1986.
3.  Everett, L.G., Hoytman, E.W., McMillan,  L.G., "Constraints  and
   Categories of Vadose Zone Monitoring Devices, Ground Water Moni-
   toring Rev. 4. 1984, 26-32.
4.  Spittler, T.M., Fitch, L.. and Clifford, S., "A New Method  for Detec-
   tion of Organic Vapors in the Vadose Zone," Proc oftheNWWA Con-
  ference on  Characterization and Monitoring of the Vadose (l/nsatu-
   rated) Zone, National Water Well Association, Worthington, OH, 1985,
   236-246.
5.  Thompson, G.M., and Marrin, D.L., "Soil Gas Contaminant Investi-
   gations: A  Dynamic Approach," Ground Water Monitoring Rev. 7,
   No. 2, 88-93.
6.  Mackey, D., and Shiu, W.Y., "A Critical Review of Henry's Law
   Constants for Chemicals of Environmental Interest. J Physical Chemistry
   Reference Data. 10. 1981, 1175-1199.
100    FIELD SCREENING

-------
                        Assessing  the  Validity  of  Field  Screening
                   Of Soil  Samples for  Preliminary Determination
                                Of  Hydrocarbon Contamination
                                                   Philip  G. Smith
                                            Harding Lawson Associates
                                                 Novato, California
                                                 Stephen L. Jensen
                                             Curtis & Tompkins,  Ltd.
                                             San Francisco, California
ABSTRACT
  A common field method for assessing soil that is potentially con-
taminated with hydrocarbons is to screen soil samples with various
gas and vapor monitoring devices. The results of the field screening
frequently are used to determine whether soil samples should be
preserved for  laboratory analysis. The results also are used to
provide preliminary estimates of the lateral and vertical extent of
contamination.
  The validity of this methodology is examined by comparing the
results obtained using various field screening  methods with the
results obtained from several laboratory analyses performed on
the same soil samples. Two field instruments and three laboratory
techniques were used to determine Total Petroleum Hydrocarbons
(TPH). A comparison of the results suggests that the field pro-
cedures  misrepresent actual contamination  as indicated  by
laboratory, derived TPH  concentrations. Linear regression analysis
of vapor concentrations measured with  gas monitoring devices
versus laboratory-derived TPH concentrations of field-screened
soils indicates that there is little statistical correlation between field
and laboratory  values.  Therefore, caution should be used in
estimating soil contamination from the results  of field screening;
field screening should not be used as the sole criterion for deter-
mining the presence and  extent  of petroleum  hydrocarbon
contamination.

INTRODUCTION
  Soil samples from contaminated and potentially contaminated
sites frequently are screened in the field with gas monitoring devices
as a preliminary estimate  of contamination. The results of the field
screening are used to determine the need for additional drilling and
sampling, to select samples for laboratory analysis and, less fre-
quently, to estimate the extent of contamination during site
remediation. In the past,  the results obtained in the field have been
assumed to be of a high enough quality to warrant their  use in
the decision-making process. If field screening does not produce
results that represent true conditions, then decisions based on these
results are, at best, suspect and, at worst, invalid. Field screening
results of hydrocarbon-contaminated soils at three sites in Northern
California were compared with the results of laboratory analyses
of the same soil samples to examine the validity of the field
screening procedure.
INVESTIGATION OVERVIEW
  The purpose of the field investigation was to obtain a preliminary
evaluation of the occurrence and distribution of hydrocarbons in
the subsurface environment.  With the exception  of the presence
of petroleum-derived hydrocarbons,  these site were considered
relatively clean. Additional volatile compounds were not detected
in the course of the investigation.
  The first two sites are petroleum products refining, storage and
handling facilities, and the third is a motor fuel storage and dis-
pensing facility.  Each has been operational for at least 15 years.
  Field screening of samples using portable monitoring devices and
visual observation was performed to determine which soil samples
would be presented to the laboratory for TPH analysis. There are
significant differences between the analytical methods performed
by each laboratory for total petroleum hydrocarbon analysis at
these three sites. Only one laboratory method was used at  each
site. The range of hydrocarbons evaluated by the laboratory tech-
nique differs between sites because of the types of extractants
(methanol, acetone and Freon 113), the methods of sample prepara-
tion (purge and trap vs. direct extraction) and quantification techni-
ques (gas chromatography or infrared spectrophotometry).
     150
I
3
                                       Numbw ol Dili PoInU . 80

                                       R . 0.31
            *


         ft .'
        D     SO     100     ISO    200     250

                    LAB TPH (Rig/Kg)


                        Figure 1
         Site One Field vs. Laboratory TPH Scattergram
Site 1
  Site 1 is a petroleum hydrocarbon storage and refining facility
that handled hydrocarbons ranging from crude oil to light-end
gasoline constituents. The facility, the oldest of the three, has been
                                                                                            FIELD SCREENING    101

-------
operating for over 25 years. It is believed that soil contamination
by petroleum hydrocarbons stemmed from surface spills, leaks in
subsurface product transfer lines and land disposal of material used
in the refining process.
   The TIP Photoionization Detector was used as field instrumen-
tation at this site. The laboratory procedure used  for TPH analysis
was U.S. EPA Method 418.1. This technique extracts a given quan-
tity of soil with Freon  113. The resulting extractant is then sub-
jected to infrared spectrophotometry for quantification of TPH
by measuring absorbance from the carbon-hydrocarbon bond
stretch. This technique can quantify a broad range of molecular
weight hydrocarbon species because Freon 113  is a fairly strong
extractant and the infrared quantification technique should detect
all hydrocarbons solubilized in  the freon matrix. Diesel fuel was
used  as the standard for laboratory analysis. Eighty laboratory
results were correlated with TPH measurements  taken in the field
(Figure 1).


Site 2
   Site 2,  like Site  1, was  a storage and handling facility  for
hydrocarbons  ranging from crude oil to lighter refined products.
The facility was operational for at  least 15 years. Leaky product
delivery lines and surface spills were suspected of contributing to
soil contamination at this site.
   Soil screening in the field was performed using  a Century Model
128 Organic Vapor Analyzer. U.S. EPA Method 3550 (sonifica-
tion  extraction with acetone  followed by  an analytical  method
similar to U.S. EPA Method 8015) was used by the laboratory for
TPH analyses. This method was recommended  for high-boiling,
point hydrocarbons by the California  Regional Water  Quality
Control  Board, San  Francisco  Bay  Region  (Guidelines  for
Addressing Fuel Leaks, September  1985). The method uses a gas
chromatograph equipped with a gas capillary column and a flame
ionization detector. Diesel motor fuel was used as a standard.
   This analytical technique is designed to quantify moderately high-
boiling-point  hydrocarbons.  However,   many of  the  larger
molecular weight species may not be detected because they may
not be solubilized  by acetone extraction and/or they may have
significantly higher boiling  points (in excess   of 350 ฐC).  One
hundred sixty-five laboratory results were correlated with the TPH
measurements taken in the field (Figure 2).
i
                                           Hunt* of Dm Potntt . 165
                                            R.OS2
                                          \
                                         Upptr Limn ol OVA
         0        1000     2000      3000      4000

                     LAB TPH (mg/Kg)


                           Figure 2
         Site Two Field vs. Laboratory TPH Scattergram
Site 3
  The investigation at this site was performed to evaluate relatively
recent product loss from defective subsurface storage tanks. Both
gasoline and motor fuels were believed to have leaked into the sub-
surface. The refined hydrocarbons present in the subsurface would
be expected to have lower boiling points on average than the broad
range of hydrocarbons observed in investigations at Sites 1 and
2. The TIP was used to field screen soil samples. Laboratory TPH
analyses were similar to  those performed for Site 2. Forty-nine
laboratory results were correlated with TPH measurements taken
in the  field (Figure 3).
                                            Numtaf of Dill Portl . 4ซ
                                            R.0.2*
                                             \
                                                   Uppw LKTW d PIO
                          1000


                     LAB TPM (mg/Kfi)
                                            2000
                            Figure 3
          Site Three FieW vs. Laboratory TPH Scattergrara
FIELD INSTRUMENTATION
  A Century Model  128  Organic Vapor Analyzer and a  TIP
Photoionization Detector  were used to measure TPH concen-
trations in  the field. The  Century  OVA is  a portable gas
chromatograph equipped with a flame ionization detector. For field
operation during these investigations, the constant sampling mode
was used. In this mode, the sample stream is constantly pumped
through the device at a fixed flow rate.  Unlike a laboratory
chromatograph, the portable  OVA does not have an oven and,
therefore, operates at ambient temperatures.
  The TIP is a hand-held device equipped with a photoionization
detector that also constantly  pumps an air sample  through the
detector. Volatile constituents entering the device are detected if
the ionization potential falls within the limits of the detector. The
ionization strength of the detector is 10.6 Ev, which enables the
detector to  respond  to a  broad  range of hydrocarbons.  Both
instruments  were calibrated in accordance with  manufacturer's
instructions  and were checked daily in the field.

SAMPLE COLLECTION
  Soil samples were obtained using a hollow stem auger equipped
with a California Modified Sampler lined with  2.5-in. diameter
brass tubes.  Soil samples to be analyzed in the field were placed
in mason jars, which were then sealed with aluminum foil. After
the sample was allowed to  equilibrate for about 5 min, the probe
tip of the field instrument was inserted through the aluminum foil
to obtain a reading. Samples for laboratory analyses were chosen
based on field screening and visual observation of contamination
(such as discoloration of soil) samples. The samples  intended for
laboratory analyses were sealed,  labeled and placed on ice for
transport to the analytical laboratory.

DATA EVALUATION
  To compare field and laboratory results, field measurements and
results from laboratory TPH analyses were plotted on a  scat-
tergram. The correlation coefficient was then calculated for each
site. Data for each investigation are presented in Figures 1  through
3. The linear correlation coefficients for these sites are: 0.31 at Site
102    FIELD SCREENING

-------
1, 0.62 at Site 2 and 0.29 at Site 3. There were several cases where
the upper detection limit of the field instrumentation was equaled
or exceeded.
  In computing the correlation coefficient at each site, the field
measurements were considered equal to the upper detection limit
value when  the laboratory TPH value was equal to or less than
the upper detection limit of the field instrument. Field TPH and
subsequent laboratory TPH values were omitted from the calcu-
lation when  the laboratory TPH value exceeded that of the upper
detection limit of the field instrument.

DISCUSSION
  The liner correlation coefficients calculated for these sites indicate
that the field and laboratory data do not correlate well. Because
the field equipment was believed to be performing correctly, the
variability between field and laboratory data can be attributed to
these major factors:

• Types of hydrocarbons in the subsurface
• Type of field screening technique
• Type of laboratory procedure for TPH analysis

  The greatest deviation between  field and  laboratory  data
probably  occurs  because  the  majority of the hydrocarbons
encountered in the soil samples are not volatile. Many gasoline con-
stituents are volatile; however, most components of diesel and
heavier fuel  oil  are  not.  Examination  of the  laboratory
chromatographs reveals that  many, if not the majority, of the
hydrocarbons  from these soil samples eluted from the analytical
columns at temperatures exceeding 200 ฐC. These higher-boiling-
point  hydrocarbons  would  not  be volatile at  temperatures
encountered in the field. Many of the hydrocarbons had  been
residing in the subsurface for an extended period of time, during
which substantial quantities of the volatile and most mobile con-
stituents may have  volatilized and/or  migrated  offsite. Con-
sequently, the resulting hydrocarbon population would be biased
toward the larger molecular weight species.
  Considering the effect of volatility, there  should be  a greater'
correlation between the field data and  the laboratory data at
Site 3 because the average molecular weight of the hydrocarbons
was assumed to be significantly less at Site 3 than at the other two
sites. Therefore, the hydrocarbons at Site 3 would be more volatile
and would be  detected with the field screening device. However,
the correlation coefficient for this site is the lowest of the three
sites. This may be attributed to the smaller sample population or
may indicate  that  the  hydrocarbons have 'aged' more  than
anticipated.
  The type of field equipment chosen to perform the screening
and the field screening technique itself are also major variables.
The screening technique used in these investigations is heavily biased
toward moderate to highly volatile constituents. Both instruments
used in these evaluations are designed to detect volatile constituents
and to respond  to  a broad  range of  compounds. Ambient
temperature, soil  type, soil sample size, container size, and time
allotted for sample volatilization  all  will affect the degree of
volatilization and, therefore, the amount of hydrocarbon sub-
sequently detected.
  The specific TPH analytical method used by the laboratory also
will add to  the dissimilarity between field and laboratory derived
results. Although TPH stands for "total petroleum hydrocarbons,"
few, if any, of the TPH procedures actually are capable of
measuring all the  hydrocarbons present in soil. In addition to the
variables associated  with analytical method, there are several
variables that affect analytical performance. The two main factors
are the strength of the extractant used and the ability of the tech-
nique to quantify large molecular weight species. The strength of
the extractant  will determine which  hydrocarbon species are
solubilized  and subsequently are available for analysis. Stronger
extractants (such  as ethyl acetate) will  effectively remove greater
quantities of large molecular weight hydrocarbons from the soil
sample than acetone. However, acetone is a more efficient extrac-
tant than methanol.
  Quantification  techniques  also play an important role in varia-
bility.  Most  quantification   techniques  based  on  gas
chromatography will not detect hydrocarbons with boiling points
in excess of 300 ฐC, which is the operational limit of most analytical
columns. The technique that  uses infrared spectrophotometry
coupled with Freon extraction will probably quantify a greater
hydrocarbon population than the gas chromatographic methods
presented above.  However, this  technique will  not  speciate
hydrocarbons. It is also more susceptible to interference from in-
digenous hydrocarbons in the soils, and it will not detect hydrocar-
bons insoluble in Freon 113. These factors indicate that the results
will vary significantly as a function of the analytical method chosen
for TPH analysis.

CONCLUSIONS
  Relying solely on field  screening results to estimate levels of
hydrocarbon contamination in the field prior to laboratory analyses
is not recommended. Several factors contribute to the variability
between laboratory and field data; the most prominent factors are
probably those of the characteristics (especially volatility) of the
hydrocarbon species present in the soil samples  and the manner
in which volatility affects field and laboratory techniques.
  Visual observation coupled with the use of field instrumentation
still may be the most cost-effective technique for field screening
soil samples for contamination by petroleum hydrocarbons, but
these should be used only as general indicators for soil contamina-
tion. Some soil samples that appear clean in the field should be
composited and tested and/or archived until preliminary laboratory
results are  known.
                                                                                                   FIELD SCREENING     103

-------
                                      Rapid  Soil  Extraction  and
                                   Cleanup  Procedure for  PCBs

                                                Jon C. Gabry,  Ph.D.
                                            Ebasco Services Incorporated
                                               Lyndhurst, New Jersey
ABSTRACT
  A rapid, soil sample extraction procedure for PCBs was deve-
loped for a mobile laboratory operating at a Superfund site. The
procedure employs sonication and includes a cleanup step using
Sep-Pak (Waters) Florisil cartridges. The new combined extrac-
tion/cleanup method was developed and validated on-site and was
approved for use on the site by U.S. EPA Region II after success-
fully passing two EMSL performance samples for PCBs. The new
procedure has a method detection limit (99% confidence interval)
of 0.6 ppm and exhibited matrix spike recoveries ranging between
82 to 105% (n = 4) for a  10 ppm of spike of Arochlor 1254 in na-
tive site soil. Additional matrix spikes at 5, 3 and 1 ppm of Arochlor
1254 had mean spike recoveries (n = 4)  of 88 ฑ 13, 87 ฑ 4 and
101  ฑ 8%, respectively. Most importantly, the new combined
extraction/cleanup procedure tripled the sample preparation out-
put of the mobile  laboratory,  without sacrificing  analytical
accuracy, to a maximum extraction output of approximately 60
soil samples per day.
  The new laboratory procedure was used to analyze 815 soil and
25 asphalt samples from the site ranging in PCB concentration from
BDL to 1000 ppm within a 6-week period. The mean matrix spike
recovery for on-site soil and asphalt  samples was 70 ฑ  20%
(n =42), whereas duplicate relative percent difference (RPD) values
for samples above 1 ppm had a mean of RPD of 14 ฑ 1% (n = 17).
Split sample analyses on  174 samples performed by CLP Labora-
tories showed good agreement with the  mobile laboratory results
(mean interlaboratory RPD for results  above 1 ppm equalled 59
ฑ  43%  (n = 50))  utilizing the  newly  developed  combined
extraction/cleanup procedure. Although not evaluated, the new
procedure also may be applicable and give acceptable results for
organochlorine pesticides as  well.

INTRODUCTION
  Within the last few years, there has been an increasing reliance
on-site mobile analytical laboratories to support super fund remedial
investigations/feasibility studies (RI/FS) and site cleanup activi-
ties. This trend has resulted from the necessity to provide a rapid
turnaround of analytical results for site characterization and/or
remedial action. Historically, mobile laboratories (or on-site instru-
mentation) initially were used solely for various field screening
applications such as determining which  samples should be sent to
an environmental laboratory for detailed analyses. In most cases,
these field analyses utilized modified methods due  to equipment
and/or space limitations which resulted  in less than optimum
accuracy and precision. Thus while acceptable for field screening,
early field results generally did not provide acceptable qualitative
                WEIGH 109 OF SAMPLE INTO A
                PRECLEANED 100 ml BEAKER
            ADD 10g ANHYDROUS SODIUM SULFATE
                     (REAGENT GRADE)
       ADD 20ml OF 1:1 PESTICIDE GRADE ACETONE/HEXANE
       AND SONICATE FOR 5 MINUTES WITH A HORNED TIP
       SONICATOR AT MAXIMUM OUTPUT. 50% DUTY CYCLE.
       IN THE PULSED MODE.
   DECANT THE SOLVENT LAYER INTO A FILTERING APPARATUS
   CONTAINING SODIUM SULFATE AND EITHER A WHATMAN IPS
   PHASE SEPARATION FILTER PAPER OR WHATMAN 41 FILTER
   PAPER .  COLLECT THE RESULTING FILTRATE (APPROXIMATELY
   10ml) IN A 60 OR 100ml GRADUATED CYLINDER .
          ADD 25ml OF 1:1 ACETONE/HEXANE SOLVENT
          TO THE SAMPLE AND REPEAT THE SONICATION
     DECANT THE SOLVENT LAYER AND COLLECT THE FILTRATE
     (APPROXIMATELY 16ml) AS IN STEP 4 ABOVE. COLLECT THE
     FILTRATE IN THE SAME GRADUATED CYLINDER USED IN
     STEP 4. BRING THE VOLUME UP TO 20ml WITH 1:1 ACETONE/
     HEXANE.
     ATTACH A SEP PAK FLORISIL CARTRIDGE TO A LUER-LOK
     SYRINGE. PRE WET THE SEP-PAK CARTRIDGE WITH APPROX-
     IMATELY 3ml OF 1:1 ACETONE/HEXANE.
     ADD 3 5ml OF SAMPLE INTO THE LUER-LOK SYRINGE AND
     INJECT THE SAMPLE THROUGH THE SEP-PAK CARTRIDGE
     AND INTO A 2ml REACTION VIAL.  SEAL IMMEDIATELY
     WITH A CRIMP CAP. ANALYZE THE EXTRACT AS PER EPA
     METHOD 8080 IN SW-846ID.
                         Figure 1
Soil Extraction/Cleanup Procedure for Polychlorinated Biphenyls (PCBs)
104    FIELD SCREENING

-------
or quantitative analytical results to support Rl/FS and/or cleanup
activities. Through a slow evolutionary process, specific equipment
and methodologies have been or currently are being developed such
that mobile laboratory accuracy and precision are comparable to
that obtained in environmental laboratories. Hence, the role of
mobile laboratories  is changing  from field  screening only to
supplying all of the environmental analyses for a site.
  Although exceptions exist, most mobile laboratories follow U.S.
EPA Method 80801 or the Contract Laboratory Program State-
ment of Work 2 for analyzing polychlorinated biphenyls (PCBs)
in soil or asphalt. Generally, a sonication extraction with a solvent
referenced in the above methods is utilized by the mobile labora-
tory. These referenced sonication extraction methods are time con-
suming since the extract generated must be dried with anhydrous
sodium sulfate and subsequently concentrated to a small volume
(approximately 1 ml). Cleanup procedures (Florisil or alumina) to
remove matrix interferences add additional steps and tune to the
extraction procedure.
  This paper presents a rapid soil extraction and cleanup proce-
dure for PCBs in soil and asphalt that was developed and validat-
ed on-site for a mobile laboratory. The method deviates from those
generally employed '-2 in that a reduced volume of solvent is uti-
lized, no concentration of the extract is performed and a cleanup
step using Florisil cartridges is used.
  By eliminating these critical time consuming steps, extraction
preparation time and equipment required are greatly reduced with
no resultant loss in precision or accuracy.

METHODS
  The development and validation of the soil extraction/cleanup
procedure for PCBs involved several steps. First, the extrac-
tion/cleanup method was proposed as a possible alternative to ex-
isting U.S. EPA approved methodology. The resultant method is
presented on Figure 1. The accuracy  of the extraction/cleanup
procedure was assessed by evaluating matrix spike recoveries from
native clean site soil. On-site soils were typically 23 to 82 percent
silt-clay (X = 55.6 ฑ 29) and poorly sorted. All method valida-
tion was performed by the on-site mobile laboratory. Extracts from
the  proposed extraction/cleanup procedure were analyzed by elec-
tron capture gas chromatography (GC/ECD) following the pro-
tocols outlined  in U.S. EPA Method 8080 '. Four matrix spikes
for  four concentrations (10, 5, 3 and 1 ppm) of Aroclor 1254 (Su-
pelco) were prepared with clean site soil. Only Arochlor 1254 mix-
ture was utilized in the method evaluation process since this was
the  only analyte of concern at the site. The extraction/cleanup
procedure method detection limit (99% confidence interval) was
then calculated from the matrix spike results to  determine if the
detection limit was acceptable for remedial action. Subsequent to
this, two Environmental Monitoring Support Laboratory (EMSL)
PCBs in soil performance evaluation samples were analyzed using
the proposed extraction/cleanup procedure. On-site soils were then
analyzed using the new method with duplicate analysis performed
at a frequency of 20% to assess reproducibility. In addition, 20%
of the field samples were split and sent to a CLP laboratory to
assess interlaboratory variability and  comparability of the new
extraction/cleanup procedure to existing U.S. EPA approved CLP
methodology 2

RESULTS
  Matrix  spike recoveries of Arochlor 1254 in native site soils
ranged between 82 to 105% (X = 91.9 ฑ 9.7; n = 4) for a 10 ppm
matrix spike utilizing the proposed extraction/cleanup procedure
(see Table 1). Additional matrix spikes at 5, 3 and 1 ppm of Aroclor
1254 exhibited mean spike recoveries (n = 4) of 88 ฑ 13, 87 ฑ 4 and
101 ฑ 8%, respectively. Based upon the 10 ppm matrix spike data,
the  proposed procedure had a method detection limit (99% confi-
dence interval) of 0.6 ppm.
  The new extraction/cleanup procedure successfully passed the
two EMSL PCBs in soil performance evaluation samples. One of
                           Table 1
     Matrix Spike Recoveries for Arochlor 1254 in Native Site Soil
               (10 PPM,  5 PPM, 3 PPM, 1 PPM)
Sanple
Designation
QA-10
QA-1)
QA-12
QA-13
QA-14
((A- 15
QA-16
QA-17
QA-18
QA-1K
QA-20
QA-21
(JA-22
QA-23
QA-24
QA-25
Sample
Arochlor 1254
Concentration
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Spiked Sample
Arochlor 1254
(ppm) Concentration (ppm)
10 ppm Spike
8.22
10.52
9.18
8.82
5 ppm Spike
4.45
3.67
4.32
5.24
3 ppm Spike
2.73
2.63
2.49
2.53
1 ppm Spike
0.98
0.94
1.00
1.12
Percent
Recover;
82.2
105.2
91.8
88.2
89.0
73.4
86.4
104.8
91.0
87.7
83.0
84.3
98.0
94.0
100.0
112.0
these samples contained Arochlor 1254 (Group II - 2.3 ppm 1254),
where the second contained Arochlor 1242 (Group 1-24 ppm
1242).  The mobile laboratory utilizing the  extraction/cleanup
procedure detected 1.4 ppm of 1254 in the Group II evaluation
sample whereas 17 ppm of 1242 was detected in the Group I per-
formance evaluation standard.
  Within a 6-week period, 815 soil and 25 asphalt samples from
the site were analyzed. Analysis was performed with two gas chro-
matographs, each equipped with two  electron capture detectors.

                           Table 2
    Matrix Spike Recoveries for Arochlor 1254 in Native Site Soil
Sample
Designation
1022
1035
1060
1079
1089
1125
1141
1146
1164
1223
1184
1204
1222
1252
Percent
Recovery
97
98
74
81
80
86
112
46
81
51
110
62
47
50
Sample
Designation
1270
1289
1309
1329
1350
1370
1390
1410
1429
1448
1468
1488
1514
1532
Percent
Recovery
56
43
54
116
66
69
55
57
52
52
46
53
95
38*
Sample
Designation
1S52
1572
1592
1612
1632
1652
1672
1692
1712
1732
1752
1771
1729
1812
Percent
Recover
72
54
45
79
62
68
75
69
74
72
106
60
77
76
  - Asphalt sample
                                                                                                    FIELD SCREENING     105

-------
                           Table 3
           Duplicate Results (Relative Percent Difference)
                For Native Site Soil Above 1 PPM


Designation
1035
1146
1223
1184
1204
1222
1289
132V
1350
• - Aspnalt

Saaiple
(ppn)
1.3
3.9
1.7
11.2
3.3
1.3
5.5
9.3
6.4
saaple
Duplicate
Result
Relative
Difference
9.3
15.5
25.0
22.0
10.0
11.0
23.0
10.0
4.0


Saanle
Designation
1468
1514
1552
1712
1732
1752
1771
1792


Duplicate
Result
Sa^le Relative
(opeO Difference
5.4 12.6
6.1 14.0
2.9 10.0
4.4 14.0
1.0 19.0*
2.0 1.0
7.9 14.0
1.2 23.0


Approximately 60 samples per day could be extracted with two
sonicators operating. This sample preparation rate with the com-
bined extraction/cleanup procedure was approximately 3  times
greater than output of the mobile laboratory following U.S. EPA
Method 8080 '. Soil and asphalt samples ranged in concentration
levels from below the detection limit (BDL) to 1000 ppm. The mean
matrix spike recovery for on-site soil and asphalt samples was 70
 ฑ 20% (n =42) (Table 2). The lowest matrix spike recovery (38^0)
occurred for an asphalt sample. This was not unexpected con-
sidering the composition of this matrix. Duplicate relative differ-

                           Table 4
                Comparison of Split Sample Results
            (CLP Versus Extraction/Cleanup Procedure)
     For Native Site Soil Containing Above 1 PPM  Arochlor 1254
Saaซ>le
Designation
S61S0106
SB37040t
SB5 10106
SBS60106
SF070206
SF030106
S002A06
SF060306
SF 100206
Sfl 70106
Sf 150406
SFI 40106
if 190206
SF3 10506
SF320306
SF200306
Sf 3 70 106
SF4602U6
M01
SF330112
SF320212
SF330312
SF3B0312
SF200412
SF110212
CIS Results-
Arochlor 12S4
Concentration
(ซ•)
1.1
4.6
19.0
20.0
4.2
7.2
4.4
1.5
2.7
2.0
6.9
3.2
1.1
87.0
150.0
S.O
60.0
6.3
0.2
39.0
18.0
92.0
3.7
16.0
7.9
function/
Cleanup fro-
cedure Results
Xrocnlor I2S4
Concentration
(MM)
t.l
2.5
15. (
1.1
2.3
12.5
1.7
1.2
1.9
2.2
3.6
3.3
2.1
59.3
M.4
4.1
21.3
3.9
1.3
12.4
36.2
17S.1
1.9
5.8
9.0
Saevle
Designation
Sf 160112
SF 130206
SF180206
SF2S05I2
ST5A-2
ST7ป-3
Sf 220106
$6220406
if 240)06
SB220I06
SF2 10606
SI2I0406
SfJ60306
Sf 360406
SF590212
SF 1301 12
SF4JOI12
SF3601 12
SF360312
SF010312
ซ. 110106
VL070106
V1050206
VL030I06
VL 090206
CLP Results-
ArocMor 1254
Concentration
(Mป)
9.7
1.9
3.6
17.0
5.6
2.4
13.0
1.0
1.8
0.9
10.0
0.3
370.0
1.6
9.0
12.0
43.0
12.0
S6.0
53.0
9.1
18.0
1300.0
65.0
32.0
Citractton/
Cleanup rro-
cedure Resulti
Arochlor 12S4
Concentration
(80.1
12.3
J.4
2.4
2.3
1.7
2.2
16.0
1.1
1.9
2.4
5.4
1.4
288.0
1.4
66.3
2.6
33.0
11.4
23.1
18.3
5.4
20.8
991.0
42.9
16.7
 ence (PPD) values for samples above 1 ppm had a mean RPD of
 14 ฑ 1% =17; Table 3). Split  sample analyses on 174 samples
 performed by CLP laboratories  showed good agreement with the
 mobile laboratory results utilizing the newly developed combined
 extraction/cleanup procedure (Table 4). Mean interlaboratory (mo-
 bile vs CLP laboratory) RPD for results above 1 ppm equalled 59
 ฑ43% (n = 50).
                                                                                             Table 5
                                                                   Comparison of CLP Versus Extraction/Cleanup Procedure Resold (PPM)
                                                                                     For Oa-ilfe Asphalt Samples
SMpl*
Designation
A4-0-1
A4-3-4
M-l-2
A6-3-4
M-2-3
O.P Results -
Arochlor 1214
Concentration (ppei)
0.16
0.69
0.16
0.16
1.20
Ex t r*c t lon/C 1 eanup
Procedure -
Arochlor 1254
Concentration (ptป)
1.30
BIX.
MIL
BDL
BDL
BDL - BeltM thซ Detection Uerit (0-4 opป)

CLP - Centric t Laboratory Progri* Laboratory
                                                                  CONCLUSION
                                                                    As the data indicate, the combined soil extraction/cleanup proce-
                                                                  dure is capable of providing acceptable analyses that are compara-
                                                                  ble to CLP laboratory results for PCBs in soil. These results were
                                                                  obtained for soils ranging in silt-clay content from 23 to 82% X
                                                                   -  55.6  ฑ  29).  Therefore, it  appears that  the  combined  ex-
                                                                  traction/cleanup  procedure would be applicable to various soil
                                                                  types. Although percent recoveries were low for asphalt samples,
                                                                  split samples sent to a CLP laboratory exhibited similar results (Ta-
                                                                  ble 5). RPDs were higher for asphalt split samples (up to 156%)
                                                                  sent to CLP laboratories; however, this is attributable to the com-
                                                                  plexity of the matrix sent (i.e., asphalt) and the fact that all quan-
                                                                  tified results (up to 1.3 ppm) were near the detection limit. Although
                                                                  not evaluated, the new combined extraction/cleanup procedure may
                                                                  also be applicable and give acceptable results for organochlorine
                                                                  pesticides as well. Several pesticides (i.e.,  Chlordane, DDT,DDE
                                                                  and possibly  Lindane) were noted in  numerous soil  samples
                                                                  analyzed  for PCBs at the site.

                                                                  ACKNOWLEDGEMENT
                                                                    Although the research described in this article has been funded
                                                                  wholly or in part by the U.S. EPA contract 68-01-7250 to Ebasco
                                                                  Services  Incorporated,  it has not been subject to the  Agency's
                                                                  review and therefore does not necessarily reflect the views of the
                                                                  Agency; no official endorsement should be inferred. Mention of
                                                                  trade names or commercial products does not constitute endorse-
                                                                  ment or recommendation for use.

                                                                   REFERENCES
                                                                   1. U.S. EPA Method 8080 in SW-846 "Test Methods for Evaluating
                                                                     Solid Waste - Physical/Chemical Methods." U.S. EPA Office of Solid
                                                                     Waste and Emergency Response, Washington, DC. 2nd Edition, July
                                                                      1982.
                                                                   2. U.S. EPA Contract Laboratory Program. "Statement of Work for
                                                                     Organic Analysis -  Multi-media, Multi-concentration." July 1985.
106    FIELD SCREENING

-------
                                   Field Analytical  Screening  for
                             Acid  Extractables  in  Soil and  Water

                                             Roger  N.  McGinn is, Ph.D.
                                                  Andrew J. Hafferty
                                           Ecology  and Environment, Inc.
                                                  Seattle, Washington
 ABSTRACT
   Ecology and Environment,  Inc., (E&E) has developed an
 analytical screening method for the analysis of acid extractable
 compounds (phenols). The method originally was developed for
 soil and water samples, but has been extended to include oil, sludge,
 wood chips and gauze wipe samples. This analytical method has
 been utilized by E&E chemists to characterize wood treating sites
 in the Pacific Northwest and to provide assistance for the emer-
 gency response cleanup of a pentachlorophenol spill.
   Sample preparation and analyses are based on modifications of
 U.S. EPA Method 604. Samples are extracted, derivatized and then
 analyzed on a mobile gas chromatograph equipped with an elec-
 tron capture detector. Disposable glassware is used where possible
 to prevent cross-contamination of samples and reduce analysis time.
 Quality control measures  include performing daily continuing
 calibrations and the analysis of method blanks, replicate samples
 and spiked samples. Correlation of data with laboratory results
 using trend analysis shows good agreement.
   The analytical screening method has been used during inspec-
 tions of four wood treating sites in the Northwest over  the past
 year. Site contamination was characterized and hot spots were iden-
 tified. Screening results were used to select samples to be sent for
 U.S. EPA Contract Laboratory Program (CLP) analysis both for
 confirmation and complete hazardous  substance list compound
 characterization. E&E has also used this  analytical method to
 provide assistance for the emergency response cleanup of a penta-
 chlorophenol/carrier oil spill.  The area of contamination was
 initially delineated using field screening. As cleanup progressed,
 analytical field screening was utilized to determine when an area
 was satisfactorily decontaminated.

INTRODUCTION
  Analytical field screening has been utilized extensively by Ecology
and Environment's (E&E's) Seattle Field Investigation Team (FIT)
for site investigations and emergency responses conducted under
the Superfund Program.
  Analytical screening which is guided by and adapted to field con-
ditions is an efficient, cost effective method to provide real-time,
site-specific data. E&E applications include:

  Detailed site characterization
  Optimization of sampling grids
  Selection of monitoring well  locations
  Selection of well screen depths
  Providing guidance to clean-up contractors
  Spill response
• Dictating remedial disposal requirements
• Identification of most appropriate samples for CLP analysis

  This paper presents an overview  of an analytical  method
developed by E&E for screening analysis of phenols and details
from several typical applications.

ANALYTICAL METHODOLOGY
  Sample preparation and analysis are based on a modification
of U.S. EPA Method 604. Phenols are extracted from soil into
methanol by vortex mixing. The sample extract is allowed to react
with pentafluorobenzylbromide and potassium carbonate in the
presence of an 18-crown-6 ether catalyst to form the pentafluoro-
benzyl (PFB) phenol derivative. The PFB derivative is exchanged
into hexane and an aliquot  of the hexane solution  is analyzed
utilizing a gas chromatograph equipped with an electron capture
detector.
  Analytical interferences are minimized through the appropriate
choice of extraction solvent and use of a selective  derivatization
reagent. Analysis with an electron capture detector also reduces
interferences since it responds only to a limited number of species.
  Method interferences may arise from contaminants in solvents,
reagents, glassware and the chromatographic system. The use of
disposable  glassware  reduces  the possibility of sample cross-
contamination while the analysis of laboratory method blanks with
each batch of samples provides a mechanism for  monitoring
interferences of contaminants which could affect analytical per-
formance.

SAMPLE PREPARATION
Soil/Sediment
  Two to three grams of sample are accurately weighed into a
disposable culture tube. Ten  ml of nanograde methanol is added
to the tube  and the sample is vortex mixed for two minutes. The
tube is then  centrifuged for 5 min, after which 5 ml of the methariol
extract are  placed into a  clean culture tube. The sample extract
aliquot is ready for derivatization.

Water
  Accurately weigh 80 - 90 g of sample into a solvent rinsed 120
ml bottle. Five drops of 50% sulfuric acid are added with  mixing.
Add 10 ml  of nanograde methylene chloride, cap  the bottle and
shake vigorously for 2 min. Allow the sample to sit undisturbed
until phase  separation occurs. The methylene chloride extraction
is repeated  twice, using 5 /d  aliquots, combining the extracts. A
solvent exchange is required  prior to derivatization. The volume
of methylene chloride is reduced to one ml under low heat and
                                                                                               FIELD SCREENING    107

-------
a gentle flow of nitrogen. The sample must not be evaporated to
dryness. Approximately 5 ml of nanograde methanol are added
to the centrifuge tube and the volume  is again reduced to 1 ml.
Final extract volume is adjusted to 10 ml with methanol. Five ml
of this methanol extract are placed into a clean culture tube. The
sample is ready for derivatization.

Sample Derivatization
  Derivative reactant  solution is prepared  by  adding one ml  of
pentafluorobenzylbromide and 1.0 g of hexaoxacyclooctadecane
(18-crown-6 ether) to  a 50 ml volumetric flask and diluting with
isopropanol. This solution remains active for one  week  when
refrigerated and stored in the dark.
  One ml of derivative solution is added to 5 ml of the methanol
sample extract in a clean culture tube. Approximately 3  mg  of
potassium carbonate are added to the tube with gentle mixing. The
culture tube is capped and placed in a hot water bath at 80ฐ C for
4 hr. After the reaction is complete, the culture tube is  removed
from the bath and allowed to cool. Five ml of nanograde hexane
are added and the mixture vortexed for one minute. Five ml  of
carbon-free water  are then added and again the mixture is vortexed
for one minute. The tube is centrifuged for  5 min after which the
hexane layer is transferred to a culture tube containing 1  g of clean
sodium sulfate. The derivatized sample, in hexane, is row ready
for analysis.

Sample Analysis
  A 2 fil aliquot of the hexane sample extract  is injected into a
gas chromatograph calibrated under isothermal operating condi-
tions shown in  Table 1.
                          Table 1
                   Instrumental Parameters
                                                                                        Table 2
                                                                            Selected Phenols-Retention Times
                                                                             1.5% S P-2250/1.95% S P-2401


                                                                                       Retention Tine (ซ1nuteป)
                                                                     C impound
                                                                2.4-DlMthylphenol
                                                                Phenol
                                                                2-Chlorophenol
                                                                2-NHrophenol
                                                                4-Chloro-3-l*ethylphenol
                                                                Z,4-01ch)orophenol
                                                                2,4,6-Trlchlorophenol
                                                                PcnUchlorophenol (PCP)
                                                                                       2m Co I urn
                          1.84
                          2.45
                          4.69
                          5.08
                          6.68
                          7.86
                          9.45
                         34.3
                                                                                                      Column
 0.52
 0.88
 1.63
 1.90
 2.34
 2.83
 3.39
12.83
                                                                                         Table 3
                                                                             Initial Calibration Detector Linearity
                                                                                 Penlachloropbenol Analysis

                                                                 Concentration (ug/1)           9ea-  Area
                                                 Relative
                                               Retention Tlw
                                                 (PCP-1.0)
                                                (2m Column)
0.054
0.072
0.137
0.148
0.192
0.229
0.27S
1.000
                                                    RF
                                                                        50                       23.174
                                                                       100                       50.251
                                                                       500                      241,639
                                                                      1,000                      547,222
                                                                      5.000                    2,608.513

                                                                       Average RF  • 0.001992             1 BSD = 6.51
                                                  0.0021S8
                                                  0.001990
                                                  0.002069
                                                  0.001827
                                                  0.001917
                            Instrument:
  Shiwdzu GC-Mint 2 Gas Chromatograph Kith electron capture detector  (ECD)

                            Integrator:

               Shinadzu Chrooatopac CR-3A Data Processor

                     Data Storage/Manipulation:

              Floppy Disk Drive, FDO-1A. CRT Display Unit

                            EC Column:
   Glass,  1 <• x 3 m. 1.51 SP-2250/1.95I SP-2401 on 100/120 Supelcoport

                           Carrier Gas:
                  95J argon/51 methane; 40 m!/minute

                   Injector/Detector Temperature:

                              275ฐC

                         Oven Temperature:
                         190ฐC. Isothernal

                         Injection Volume:
                             2.0 ul

                         GC  Analysis T1ne:
                            20 Blnutes
   Compound identification is based on retention times compared
 to those of standards. Since retention times are dependent upon
 analytical conditions, standards should be reanalyzed whenever in-
 strument changes are made. A second, dissimilar chromatographic
 column may be used, if desired, for further confirmation of com-
 pound identity. Typical retention times of a mixed phenol standard
 are illustrated  in Table 2.
                                                                Samples are quantitated using the external standard method.
                                                              Pentachlorophenol standards of known concentration are carried
                                                              through the entire derivative preparation procedure for each batch
                                                              of samples derivatized. Prior to sample analysis, an initial calibra-
                                                              tion is performed to obtain detector response factors (RF) from
                                                              the following equation:
                                                                       Concentration Std.  Injected                       (1)
                                                                            Std. Peak Area
RF =
                                                                During the initial calibration, RFs are determined by generating
                                                              a five point calibration curve to insure detector linearity. Standard
                                                              concentrations are selected which would bracket expected sample
                                                              extract concentrations. An integrator is programmed to generate
                                                              RF data prior to calibration. In order to ensure detector linearity,
                                                              the percent relative standard deviation (% RSD) for the RFs, as
                                                              calculated by the equation below, should be less than  15%.
                                                                          RF Standard Deviation
                                                                                 Mean RF

                                                        (2)
                                                                Typical detector RF linearity data from this study are presented
                                                              in Table  3.

                                                                A continuing calibration is  performed daily for each batch of
                                                              samples derivatized to ensure detector stability. A derivatized mid-
                                                              range  pentachlorophenol  standard  is  injected  into  the  gas
                                                              chromatograph and a new RF is calculated. The RF stored in the
                                                              integrator is updated with  the new value unless the percent dif-
                                                              ference between the new RF  and the average RF of the initial
                                                              calibration is  greater  than 20%.  In that case,  a  new  multi-
                                                              concentration initial calibration is performed and a new RF entered
                                                              into the integrator. Typical continuing calibration data from this
                                                              study are presented in  Table 4.
108
FIELD SCREENING

-------
                           Table 4
        Continuing Calibration Pentachlorophenol Analysis
Date
           Initial  Calibration
             Response Factor
12/13/86        0.001992
12/17/86        0.001992
12/18/86        0.001992
12/22/86        0.002172
2/26/87         0.002172
  Continuing
  Calibration
Response Factor
                                   0.002191
                                   0.002199
                                   0.002099
                                   0.002168
                                   0.001854
                                                    Difference
                      10.0*
                      10.4*
                       5.451
                       0.2%
                      14.6*
                                                        (1)
  Following instrument calibration, a 2.0 /*! aliquot of derivatized
sample extract is injected into the GC for analysis. The time
required for chromatographic analysis is approximately 20 min,
depending upon  analytical conditions, to ensure all compounds
have eluted off the column.
  Sample and standard chromatograms are printed out on the
integrator. If a peak is identified as pentachlorophenol, based on
retention time, area under the peak is used to compute the con-
centration of  solid samples by the following  equation:

                     Peak Area x  RF x Extract Volume (1)
    Cone (/4g/kg) =  	x Dilution Factor	
                     Sample Wt.  (kg)

QUALITY CONTROL  / QUALITY ASSURANCE
  Initial instrumentation set-up follows standard laboratory pro-
cedures which encompass measures to ensure a  clean, leak-free GC
system.
  Nanograde  reagents and analytical standards prepared specifi-
cally for U.S. EPA  methods are used for all analyses. Sealed
standard and spiking solutions are refrigerated at all times to main-
tain standard  integrity.
  Standard response is  monitored  and tracked.  Standards are
replaced regularly or if response factors indicate deterioration or
significant change.  Spiking solutions are periodically checked
against standards.
  All glassware is washed with laboratory detergent and rinsed with
deionized and/or carbon-free  water followed by three solvent
rinses;  methanol,  acetone and hexane. All glassware is covered and
stored  in a clean environment. Normal field conditions require the
analyst to exercise caution in order to avoid contamination and
cross-contamination  of the analytical  equipment,  glassware,
solvents, standards, samples and anything used or present in a field
screening activity.
  Initial calibration of the GC detector response is done using
mixed  standard solutions of known concentration. Calibration is
performed at a minimum of three concentrations to insure detec-
tor linearity over the range of interest. The detector response factor
(RF) for each compound is calculated as outlined  earlier. The
percent relative standard deviation (% RSD) of the response factors
for  each compound should be less than 15%.
  Continuing  calibration  is checked daily to insure detector
stability. Response factors determined using the medium concen-
tration standard solution should be within 20% of the mean RF
determined from the initial calibration.

Method Blanks
  Laboratory method blanks are run a minimum of once per day
or once every 20 samples, whichever is more frequent.  Empty
sample vials or carbon-free water are carried  through the entire
procedure for each matrix-type analyzed.

Replicate Analysis
  Replicate analyses are used to evaluate analytical precision within
a data set.
  Precision is defined as the tendency for replicate results to exhibit
grouping about a central "point." In field screening, precision is
primarily a function of sample size and homogeneity. The inherent
limitations of field screening generally preclude the analyst's
obtaining replicate samples  with  identical matrices which are
required for precision to have true statistical significance for the
analysis. Precision in field screening includes a sampling error com-
ponent  that cannot be avoided.
  Aqueous samples and dry soils  with uniform grain size yield
higher precision data since care can be taken to assure sample
homogeneity. Other indeterminate errors may be minimized by
employing good, standard analytical technique.
  Replicate sample analyses are performed, at a minimum, on 5%
of a sample set or once per batch of samples, whichever is more
frequent,  for each matrix analyzed. Duplicate analyses normally
are sufficient for water samples, while triplicate analyses often are
desirable for soil/sediment samples.
  Results for a number of  replicate analyses  are  presented  in
Table 5.
                                                                                              Table 5
                                                                                       Replicate Sample Results
                                                                                      Pentachlorophenol Analysis
                                                                         Sample I
                                                                                          Concentration
                                                                                                        Replicate
                                                                                                      Concentration
                                                                                    Relative
                                                                                   1 Difference
CSL-1
CSL-2
CSL-12
CSL-04
SOS Al
507 Al
S07 Bl
ACO 01
SOI Al
503 Bl
S03 Cl
OU-1
BN-SO-8
KN-SO-17
JHB-2
JHB-12
Soil
Soil
Soil
Soil
Soil
Soil
Soil
Water
Soil
Soil
Soil
Oil
Soil
Soil
Soil
Soil
100 U
5,800,000
100 U
60 J
SO J
3,300
100 U
1.800
170
26,000
56,000
1,470
100 U
370
75,000
380,000
100 U
2,400,000
100 U
100 U
80 J
2,300
100 U
1,400
460
25,000
32,000
1,300
100 U
360
74,800
394,000
0.01
83.0%
o.ot
o.ot
46.01
35.7S
o.ot
25. OS
92.01
4.01
55.01
12. OS
O.OS
2.7J
0.3S
3.7J
                                                                      U  - The material was analyzed for, but was not detected.  The associated numerical
                                                                         value 1s an estimated sample quantltation limit.

                                                                      J  - The associated numerical value 1s an estimated quantity because quality con-
                                                                         trol criteria were not met or concentrations reported were less than the esti-
                                                                         mated sample quantltation limit.
                                                                    Spiked Sample Analysis
                                                                      Accuracy is defined as the closeness to which analytical results
                                                                    approach the true value. Although it is  not possible to ensure
                                                                    absolute  accuracy for environmental  samples,  spiked sample
                                                                    analyses provide a measure of extraction efficiency and sensitivity
                                                                    and thus, indirectly, the accuracy.
                                                                      A minimum of one spiked sample of each matrix type is analyz-
                                                                    ed with each batch of samples. Samples are spiked with pentachlor-
                                                                    ophenol standards of known concentration dissolved in methanol.
                                                                    Results  for spiked sample analyses are  presented in Table 6.
                                                                    Recoveries typically are higher than observed by CLP laboratories.

                                                                    Laboratory Correlation
                                                                      A limited number of reportedly duplicate  samples  from a site
                                                                    were analyzed by both E&E's field screening laboratory and a com-
                                                                    mercial laboratory. Since the private laboratory is not part of U.S.
                                                                    EPA's Contract Laboratory Program, no quality assurance data
                                                                    were available to ascertain laboratory performance. However, com-
                                                                    parison of data sets and laboratory duplicate results presented in
                                                                    Table 7 can be  useful for data trend analysis. In general, field
                                                                    screening results were lower than laboratory results. Linear regres-
                                                                    sion analysis of the two data sets gave a slope of 0.795, an intercept
                                                                    of -5,600 and  a correlation coefficient of 0.91.
                                                                                                    FIELD SCREENING     109

-------
                              Table 6
                       Spiked Sample Results
                     Pentachloropbenol Analysis
                             Table 7
                      Laboratory Correlation
                     Pentacblorophenol Analysis

Suple 1
CSl-12
ACS-L01
SO? Bl
ACO 01
6N-SO-8
KN-SO-17
JHB-23

liltrll
Soil
Soil
Sod
Miter
Sotl
Soil
Soil
Spiked
Staple Result
2.500
59,000
2,100
4,800
1,350
1,800
1.980
Simple
Result
100 U
27,000
100 U
1,800
100 U
360
380
Amount
Spiked
2.000
75.000
2.000
2.000
2,500
2.500
2.000

1 Recovery
1251
izsi
1051
1551
MI
5M
801
                                                                                5topic ซiaซer
                                                                                              EtE FIT Result
                                                                                                            UMritory tesult
    U   The Mterttl wis iiulyied for, but vis not detected.  The iisocllled nunertci!
       vilue Is ซn esttnuted swple quentltltlon Unit.
   While data exhibited similar trends, agreement was lower than
 expected  for  duplicate  samples. Communications with  field
 sampling personnel indicated samples were not homogenized after
 collection, and therefore, were not true duplicates.

 CONCLUSION
   The phenol analytical screening method developed by E&E pro-
 vides  high  quality,  site-specific  analytical  data  on  a  rapid
 turnaround  basis. Screening results show good correlation with
 commercial  laboratory analysis.
   The procedure has been successfully used to delineate areas of
ACS 101
ACS SO)
SOI Al
SOI 11
SO) Al
SO) 11
SO) M
SO) Cl
SO) C4
S04 Al
S04 II
SOS Al
SOS 11
CSL-IS
CSl-22
Transfer II ink
12,000
4. MO
I/O
M
11,000
2ป,000
25.000
56.000
32.000
74,000
250.000
10.000
MO
30,000
IM.OOO
100 U
D.OOO
14, 000/2) .000 -
WO/WO
1)0
)7.000
55,000
100.000
87,000
17.000
IS, 000
MO.OOO/2M.OOO -
8,000
400
170.000
185.000
100 U

AlplluU
efcpllute







tfupllute





    Llnetr turns loo

                 Slope • O.m
              Intercept • - $.600
    Correction Coefficient • O.tl
contamination at both active and inactive wood treating sites and
to select the most appropriate samples for litigation quality CLP
analysis. Field screening has been used effectively in emergency
response cleanup to determine both the extent of contamination
and to release areas which have been adequately decontaminated.
110    FIELD SCREENING

-------
                              Computer  Modeling  Results and
                               Groundwater  Treatment System
                                   For Price's  Landfill  No.  1

                                         Abu M.Z. Alam, Sc.D., P.E.
                                         Camp Dresser & McKee Inc.
                                              Edison, New Jersey
                                                  George Klein
                           New Jersey Department of Environmental Protection
                                             Trenton, New Jersey
                                             Salvatore Badalamenti
                                                Robert McKnight
                                   U.S. Environmental  Protection Agency
                                             New York, New York
ABSTRACT
  Price's Landfill No. 1 is a 26-acre Superfund site which origi-
nally was a sand and gravel mining operation. The site began
operating as a landfill in 1968 and  began accepting bulk and
drummed liquid and solid wastes including hazardous substances
in 1971. The total quantity of hazardous substances dumped at
the site has been estimated to be about million gallons.
  To determine the extent of groundwater contamination at this
site and its  impact on potable water supplies in the area, the U.S.
EPA and the NJDEP contracted with Camp Dresser & McKee Inc.
to conduct  a Remedial Investigation/Feasibility Study; computer
modeling of the flow and contaminant transport; treatability studies
for groundwater extraction and treatment; and a conceptual design
of a remedial treatment system. This paper presents the results of
the three-dimensional finite element computer models used  to
describe groundwater flow and contaminant transport at Price's
Landfill No. 1 and the proposed treatment system  for the con-
taminated groundwater. Details of model calibration using data
collected hi the field and the basis for selection of the treatment
system for  remediation are presented.

INTRODUCTION
  Price's Landfill No. 1, is the 26-acre site of a former sand and
gravel operation located in Pleasantville City  and  Egg Harbor
Township in New Jersey. Sand and gravel excavation operations
at this site ceased during 1966, and limited amounts of construc-
tion wastes were accepted at the site as fill materials for open
excavations. In 1969, the  Price's Landfill No. 1 site became a
commercial solid waste landfill.
  In May 1971, Price's Landfill No. 1 began to accept a combina-
tion of bulk and drummed liquid wastes. Available information
indicates that these wastes included industrial chemicals, sludges,
oils, greases, septage and sewer wastes; some wastes were dumped
in bulk from tanker trucks, while  others were buried in 55-gal
drums. Information available also indicates that some of the drums
were punctured prior to burial. Total quantities  dumped are
estimated at approximately million gallons. Chemical waste disposal
operations  were terminated in November 1972. Sludge disposal
continued until May 1973 and municipal waste disposal was ter-
minated in 1976.
  Price's Landfill No. 1 was ranked number six on the June 1984
U.S. EPA NPL. In 1980, residential wells in the area were found
to be contaminated with volatile organic compounds. Since then
the New Jersey Department of Environmental Protection (NJDEP)
and the U.S. EPA have undertaken a number of steps in prepara-
tion to mitigate groundwater contamination in the area adjacent
to Price's Landfill No. 1. In December 1981,37 affected residences
were connected to the New Jersey Water Company (NJVC) water
supply system. From January 1982 through May 1983, a RI/FS
was undertaken by the U.S. EPA at the site.
  During 1984 and 1985, under a Cooperative Agreement between
NJDEP and U.S. EPA, Camp Dresser & McKee Inc. (COM) con-
ducted a second RI/FS at Price's Landfill No. 1.  The results of
this RI/FS form the basis of this paper.
  During  the 1984 Remedial Investigation at Price's Landfill
No. 1, it was determined  that the site and the surrounding area
were subject to groundwater contamination. Results obtained from
sampling the groundwater  indicate that a significant source of con-
tamination remains in the landfill since the groundwater down-
gradient from the site was highly contaminated with chemicals
dumped on-site. Sampling and analysis of the groundwater
indicated  the presence  of benzene,  cadmium,  chloroform,
dichlorethylene, lead,  1-2-trans-dichloroethylene, trichlorethylene,
vinyl chloride and acetone  in the upper Cohansey formation. Total
organic volatiles (TOY) ranged from 40 to 50 jig/1 near the land-
fill in the shallow depths of the upper Cohansey. TOV levels ranged
from 0.01 to 1 /tg/1 in the deeper areas of the aquifer.

COMPUTER MODELS USED
  Two  computer  models,  DYNFLOV and DYNTRACK,
developed by CDM were used to simulate the groundwater flow
                              ACMUA WEU-S
                              DEP WELLS
                         Figure 1
           Price's Landfill No. 1 and Detailed Features

           MODELING OF HAZ MAT TRANSPORT    111

-------
and contaminant transport processes to determine the likely extent
of contaminant migration from the landfill and to estimate the ef-
fectiveness of the various remedial  schemes.
  DYNFLOV is a fully three-dimensional finite element model for
groundwater flow that can be used to simulate steady-state and
transient responses of groundwater  flow systems to several types
of natural and artificial stresses. The program employs linear finite
elements and  can simulate induced infiltration from  streams,
artificial and natural recharge or discharge, and nonhomogeneous
and anisotropic aquifer hydraulic properties.
  DYNTRACK was  used to simulate the paths of contaminant
migration from the landfill under various proposed remedial alter-
natives. DYNTRACK can perform either single particle tracking
or can simulate three-dimensional contaminant transport of con-
servative or first order decay constituents while  allowing for
dispersion and retardation. The mass transport section of the code
utilizes the "random-walk"  method for statistically significant
numbers of particles, with each particle having an associated mass
and decay rate. Both  DYNFLOW and DYNTRACK use the same
finite element grid. DYNTRACK uses the flow field generated by
DYNFLOV to simulate contaminant transport.

CONCEPTUAL MODEL

Surface Features
  Major surface waters in the vicinity of the landfill include the
Atlantic City Municipal Utilities Authority (ACMUA) reservoirs,
Absecon Creek and Conover Run. Two major rivers, the Mullica
and the Great Egg Harbor, form a hydrologic border to the region
of the county in which Price's Landfill No. 1 is located. Surface
water and groundwater near the landfill flow in a generally eastern
direction toward the Atlantic Ocean. Conover Run,  a tributary of
Absecon Creek, receives some of this groundwater flow.

Site Geology
  The principal freshwater aquifers in Atlantic County lie in the
Kirkwood Formation and the Cohansey Sands. The Kirkwood
Formation consists chiefly of sand, silt and clay and  varies in
thickness from 100 ft. at the northeastern border of the county
to 700 ft.  at  Atlantic City. Near the landfill, the Kirkwood
Formation is approximately 600 ft. thick. Elevation of the top of
the Kirkwood at this location is about 225 ft. below mean sea level
(MSL), or about  260 ft. below the ground surface.
  The Cohansey Sand aquifers conformably overlie the Kirkwood
throughout the region near Price's Landfill No. 1. The Cohansey
Sand is an unconsolidated deposit of quartz sand that contains some
gravel and notable amounts of clay.  One such major clay bed with
thicknesses between 20 and SO ft.  divides the Cohansey Sand into
two major units: the upper Cohansey and the lower Cohansey. This
clay layer has  been detected  in all areas near the landfill and in
the area of the  new ACMUA well field north of the reservoir. Based
on  review of  well logs and bore hole  gamma logs, the upper
Cohansey sand was viewed as two distinct zones separated by an
intermediate clay zone. This clay zone acts as an aquitard to main-
tain a piezometric head difference of about 3 ft. across the clay
layer.

Extent of Site Contamination
  The water quality data collected at the site were reviewed in order
to determine the location and extent of contamination. This data
review indicated that groundwater contamination extends from the
Price's Landfill No. 1 to about 1 mile east of the landfill (beyond
New  Road), with  the bulk of the  "plume"  following an east-
northeast to a northeast direction towards the intersection of
Conover Run  and Absecon Creek.
  The surface unit of the  upper Cohansey  aquifer shows the
presence of high levels of contamination immediately downgradient
of Price's Landfill No. 1. However, the levels of contamination
decrease rapidly in this unit within a short distance from the landfill.
  The lower zone of the upper  Cohansey aquifer exhibits the
presence of a contaminant  plume which stretches from  just
downgradient of the landfill all the way to the monitoring wells
located in the vicinity of New Road.

Model Geometry
  Following a detailed review of the stratigraphy and the ground-
water quality data from the site, it was decided to use a five layer
(six level) representation of the stratigraphy for the DYNFLOV
and DYNTRACK models of the area. Figure 2 depicts the general
vertical configuration of the  model.
                          Figure 2
 Vertical Cross-section and Model Layers Used at Price's Landfill No.
  Each layer in the model corresponds to one unit of the Cohansey
Sand described earlier. As shown in Figure 2, which represents a
cross section of the model running from the landfill to the north-
east toward Absecon Creek, level 6 represents the ground surface
and the top of the upper Cohansey formation. Layer 5 corresponds
to that portion of the aquifer lying above the zone of clay lenses
discovered at the 50- to 80-foot level. These upper Cohansey clay
lenses are represented by layer 4. Layer 3 represents the lower por-
tion of the upper Cohansey aquifer, which  is the zone primarily
used for water supply.
  Layer 2 represents the mid-Cohansey clay. This layer generally
varies in thickness between 5 and 40 ft. throughout the modeled
area. Near the landfill and in the area impacted by the plume, the
thickness is approximately 20  ft. as estimated from boring logs.
Layer 1, the bottom layer, represents the lower Cohansey unit and
constitutes the bottom  of the model representation. It is estimated
that a dark  gray-green clay layer  located within the Kirkwood
Formation immediately under the lower Cohansey acts as a signifi-
cant barrier to communication with the water-bearing units within
the Kirkwood Formation since it is generally over 25  ft. thick.
  The horizontal limits of the  model were extended, as shown by
the grid in figure 3, to be sufficiently distant from the site to
minimize their  influence on  the flow field at the  site and its
immediate environs.
  The region modeled  is about 47 sq. miles in area. The horizontal
finite element grid used for simulation contains 337 nodes in each
of the six levels for a total of 2,022 nodes. It also has 636 elements
in each of the five layers for a system-wide total  of 3,180 elements.
More detail with a higher density of nodes and  elements was pro-
vided in the area of greatest interest immediately downgradient of
the landfill.
Boundary Conditions
  Specified (fixed) head boundary conditions were applied to all
six levels of the perimeter nodes in the model.  These heads were
based either on field data or,  in some cases, on heads computed
112    MODELING OF HAZ MAT TRANSPORT

-------
by a regional model which covered a much larger area. To repre-
sent the interior water bodies, specified heads were assigned to level
6 at average water elevations. Absecon Creek and its branches were
simulated using the rising water feature of the DYNFLOW model.
Recharge
  Recharge from rainfall was applied at all ground surface (level
6) nodes in the model. Based on the hydrologic budget of the area,
an applied recharge value of 20 in. per year was used. The resulting
average effective recharge over the total modeled area was estimated
to be about 15.7 in. per year.

Model Parameter Values
  For each element in the model, the horizontal and vertical
hydraulic conductivities were needed.  An initial set of aquifer
parameter  values was estimated and  used to perform the first
calibration runs of the steady-state flow field. During the calibra-
tion stage,  the values of the hydraulic properties for the elements
were varied over the ranges shown in Table 1. The final adopted
set of parameter values also is  shown  in this table.

                           Table 1
                  Aquifer Hydraulic Properties

Horizontal Hydraulic Conductivity
(KxxandKyy)
Upper Cohansey Sand - Layer 5
Upper Cohansey Sand - Layer 3
Lower Cohansey Sand - Layer 1
Upper Cohansey Clay -Layer 4
Mid Cohansey Clay - Layer 2
Vertical Hydraulic Conductivity
(Kzz)
Upper Cohansey Sand - Layer 5
Upper Cohansey Sand - Layer 3
Lower Cohansey Sand - Layer 1
Upper Cohansey Clay - Layer 4
Mid Cohansey Clay - Layer 2
Initial
(tl/day)


60
60
60
0.0226
0.0226


6
6
6
0.000226
0.000226
Range Tested
(It/day)


25-150
45-150
45-240
0.0226-20
0.0226-0.2835


22-15
45-15
45-24
0.000226-0.2635
0.000226-0.0284
Final
(fl/day)


40
80
140
0.2268
0.09


4
6
14
0.075
0.0072
                                         •"V.
• />
-------
                            Table 4
          Observed and Calibrated Plezometric Head Value*
                 For Upper Cohansey Formation
                           (Layer 1)


                        U Vltltl M FMI (MSI)

        WHIG       MtnutMHMCl    CofflpulKl Htld     Dซซrซncซ
       A-1
       A3

       C-4A
       C-SA

       P-2
       P-3
       P-5
       DEP-040
       OEP-oao
       DEP-100
       OEP-110
       D6P-12D
       0€P-IJO
(21
•i.21

• 42
•ซ32

-0.2ซ
-014
072
•202

• 25
-447
•110
•277
•137
100
2.32
•745
-0.55
•041
-0ซป
US

4.72
-331
-341
-010
1 55
446
•111
071
309

•0.21
-027
•1 40
OM

-153
101
!M
1.17
It2
-352
       MEAN DIFFERENCE -010
       STANDARD DEVIATION 2-33
source points initially was set to a uniform value. These strengths
were varied during the calibration process so that the concentra-
tions in the computed plume better matched the Held measured
data.
                                  ACMJAWtuS
                                  DEP WELLS
                                  A WEILS
ซ:   '''.'•-.  •?    '••'   /   '
*:'   /  />'•-•   /  /
 !.  /  /'    "•/-..  '
                                                                                                Figure 5
                                                                           Simulated Contaminant Concentrations in Model Layer S
                                                                                                 Figure 6
                                                                            Simulated Contaminant Concentrations in Model Layer 4
                            Figure 4
          Contaminant Plume Movement and Area Impacted
                     At Price's Landfill No.  1
CALIBRATION RESULTS
  The calibrated contaminant plume for present conditions was
developed by simulating the 1971 through 1984 period. The overall
movement and total area impacted by the plume are illustrated in
Figure 4, where the plume's generally easterly migration is shown
by the distribution of the "particles." The concentrations of total
volatile organics in each of the top four layers of the model are
shown in Figures 5 through 8, while Figure 9 illustrates the plume
cross section along its centerline.
  The  computed  concentrations  at  various  monitoring well
locations were compared to the observed values of total volatile
organics in the wells sampled during the July 1984 sampling round.
Tables 5 through 7 show the comparison between the field measured
data and the computed concentrations at each  of these wells. As
seen from the results, the model predictions provide a reasonable
representation of the contaminant distribution in the field.
  As seen in Figures 7 and 9,  the simulated  plume extends to
beyond New  Road in the lower zone of the  upper Cohansey.
                                                                            Figure 7
                                                       Simulated Contaminant Concentrations in Model Layer 3
114    MODELING OF HAZ MAT TRANSPORT

-------
                                                                                             Table 5
                                                                    Measured and Computed Concentrations of Total Volatile Organics
                                                                                        At Monitoring Wells
                           Figure 8
      Simulation Contaminant Concentrations in Model Layer 2
MonHoring Well
Upper Cohansey

C-2C
C-3C
C-4D
C-5C
EPA-1
EPA-1 A
EPA-2
EPA-3
EPA-4
EPA-5
EPA-5A
DEP-1
DEP-2
OEP-3
DEP-4S.
OEP-5.
OEP-6*
DEP-7ป
DEP-9S
OEP-101
DEP-12S
OEP-14S
DEP-1 SS
1964 Measured
(Ranoe)
(PPb)
ND-2
ND
ND-1.3
ND-6
5664-10203
12243-21800
ND-5
ND-10
ND-10
NA
NA
51610-76640
NA
16-25
NO- 11
17420-24220
10605-12910
4305-5706
13617-16740
150-160
22-27
64-244
204-249
Previous
Measurements
(ppbl
ND-13
ND-15
3-33
2-300
500-37,000
64-2,175.000
ND-393
ND-Bt
ND-41
ND-132
ND-226
ND-1 40.000
16,456-71,000
NO-12E









Computer
Results
(Ranoel
(PPb)
ND
NO
NO
ND
5997
369-3,464
ND
ND
ND
ND
NO
3,950
269.000
0-126
ND
15,564
23.667
5,373
59.364
ND
ND
3.127
ND-600
                                                                           NOTE: Wets are screened In Melton ol layer 4 (Upper Cohansey Clay).

                                                                                             Table 6
                                                                    Measured and Computed Concentrations of Total Volatile Organics
                                                                                        At Monitoring Wells
                           Figure 9
    Vertical Distribution of Contaminants Along Plume Centerline
Concentrations in this  layer range from about  100 /tg/1  just
downgradient of the landfill to about  10 jtg/1 at New  Road.
Contamination in the upper layers is largely confined to an area
just beneath and downgradient of the landfill where  concentra-
tions in the range of 10-50 mg/1 have been observed.
  This predicted downward movement  is further illustrated in
Figure 9, which shows a cross-sectional view through  the plume.
This figure shows the general behavior observed in the contami-
nant plume, with the plume moving quickly down through the top
layer of the upper Cohansey sand and through the upper Cohansey
clay to the lower zone of the upper Cohansey formation. The con-
tamination is then generally predicted to move horizontally in the
formation with some downward movement through the vertical
extent of this layer. Some slight penetration into the mid-Cohansey
clay layer also was predicted. This penetration markedly increases
in areas with overly active production  wells where  large head
gradients exist across this clay zone.

SIMULATION OF ALTERNATIVES
  Following calibration, the DYNFLOV and DYNTRACK models
were used to simulate the effectiveness of  the remedial alternatives
for mitigating plume contamination.  In each case, the alternatives
Monitoring Well
Upper Cohansey

A-4
C-1A
C-2A
C-2B
C-3A
C-3B
C-4B
C-4C
C-5B
C-6
C-7
C-8
C-9
DEP-8I
DEP-9I
DEP-1 11
DEP-121
DEP-13S/1
DEP-41
P-1
P-4
P-7
P-8
P-9
P-12
1964 Measured
(Haryje)
(PPO)
2.6-29
7-9.7
ND-1. 2
ND-1.2
ND
ND-27
ND-1.2
ND-1.2
27-28
4.094-7.672
ND
13.5-33
ND
ND
656-1632
NO
15-17
ND
NO
ND-1
NA
NA
1.3-1.2
NA
NA
Previous
Measurements
(PPdl
ND
40-115
ND-6
ND-7
ND-1 30
6-516
ND
3-26
11-22
ND-799
10-177
3-306
ND






ND-35
NA
7
NO-57
24
NA
Computer
Results
(Range!
(PPb)
ND
ND-101
15
31-6
ND
ND
ND
ND
3.6-21
ND
ND
9
ND
ND
657
11
11
ND
142
ND
ND
65
102
ND
6
     NOTE: DEP-91 and -121 are screened partially m this formation and partially In the day layer table

                         Table 7
Measured and Computed Concentrations of Total Volatile Organics
                   At Monitoring Wells

Monitoring Well
Lower Cohansey
A-1
A-3
C-4A
C-5A
P-2
P-3
P-5
P-6
DEP-4D
OEP-8D
DEP-10D
DEP-11D
DEP-12D
DEP-13D

1964 Measured
IRanae)

-------
were evaluated over a time span of 20 years from 1984, under the
assumption  that Price's  Landfill No.  1  would continue as a
contaminant source for at least that period.
  Steady-state flow fields from DYNFLOV were used as the in-
put to DYNTRACK. When existing water supply pumping changes
to a new pumping scheme, the resulting new steady-state flow field
was used. The model simulations are based upon the assumption
that Price's Landfill No. 1 is the only source of groundwater con-
tamination in the area being studied. Three remedial alternatives
were analyzed in detail using the DYNFLOV and DYNTRACK
models. These are briefly described below.

Alternative  1 - No Action.
  This alternative calls for no remedial action at the site. The only
action was the installation of nine new production wells in the lower
Cohansey aquifer  to the north of the Atlantic City reservoir to
replace the existing ACMUA production wells in the upper and
lower Cohansey formations that are east and northeast of Price's
Landfill  No. 1. The key features of the resulting plume after 20
years are shown in  Figures 10 and 11. These figures show that both
zones of the upper Cohansey sands will remain contaminated, as
will the upper Cohansey clay. Some movement of contaminant into
the mid-Cohansey clay also is possible, but contamination is not
expected  to penetrate  the lower Cohansey formation itself.

Alternative 2 - Plume Abatement Pumping
  This alternative also presumes the replacement of all existing
ACMUA Cohansey Formation wells with the new wells in the lower
Cohansey north of the reservoir. In addition, plume extraction wells
would be installed immediately to the east of the landfill and the
landfill area would be  capped. The water pumped from  the
extraction wells would be treated and discharged in an approved
manner. The predicted  control impacts of this alternative are clearly
shown in Figures 12 and  13 where plan views of the particle loca-
tions at years 5 and 20 are shown. The plume immediately down-
gradient of the landfill is largely controlled by year 5. Some con-
tamination, which is already downgradient of the pumping system,
will continue to move to the east.
                          Figure 10
      Contaminant Concentrations in Model Layer 5 at 20 Years
                  Under No-Action Alternative
                           Figure 12
       Contaminant Concentration in Model Layer 5 at 20 Years
          Under Plume Pumping and Treatment Alternative
                                          ttป!ป 1000*1 (Mf
                           Figure 11
       Contaminant Concentrations in Model Layer 3 at 20 Years
                   Under No-Action Alternative
                                                                                                             ซ•>•; 10001 |M<
                           Figure 13
        Contaminant Concentration in Model Layer 3 at 5 Years
           Under Plume Pumping and Treatment Alternative
 116     MODELING OF HAZ MAT TRANSPORT

-------
Alternative 3 - Containment Vail with
Plume Abatement Pumping
  This alternative consists of an 80-ft-deep containment wall which
surrounds the landfill and which is designed to contain any source
material remaining within the landfill. The landfill area would be
capped at the time of installation of the wall. In addition to the
installation of the containment wall, this alternative includes the
plume control pumping similar to that in Alternative 2, together
with pumping from within the containment wall itself. The results
of this extensive control of both the source and plume are shown
in Figures 14 and 15, where the predicted concentrations in layer
5 and  3 at 5 years are shown.
  lEQEHD
                                           seal*: IDOO'i (••
                          Figure 14
       Contaminant Concentration in Model Layer 5 at 5 Years
    Under Slurry Wall, Plume Pumping and Treatment Alternative
 v	Cr**ht
                           Figure 15
        Contaminant Concentration in Model Layer 3 at 5 Years
     Under Slurry Wall, Plume Pumping and Treatment Alternative
RECOMMENDED ALTERNATIVE
GROUNDWATER EXTRACTION
  Based on the results of computer modeling, the groundwater
extraction program designed to meet site remediation objectives
was recommended for Price's Landfill No. 1. Groundwater will
be extracted from the upper portion of upper Cohansey at a rate
of about 200,000 gal/day. Groundwater from the lower portion
of the upper Cohansey will be extracted at a rate of about 1.1 mil-
lion gal/day.  In order to restore groundwater quality and abate
the contaminant plume, it is anticipated that groundwater extrac-
tion from the upper portion of the upper Cohansey will be required
for a period of approximately 25 years. It is expected that pumping
of the lower portion of the upper Cohansey will be necessary for
a period of approximately 5 years.
  Results of computer modeling indicated that at least three shallow
wells would be required  in the upper Cohansey to create a large
enough cone of influence to control migration of the contaminant
plume. The preliminary locations of these wells are shown in
Figure 16. The exact number, capacity and location of each shallow
extraction well will be finalized based on the results of pump tests
to be conducted during  final design.
  The computer analysis indicated that a minimum of two deep
wells should be provided to extract  1.1 million gal/day of ground-
water from the  lower  portion of the upper Cohansey.  The
preliminary locations of these wells are shown in Figure  16. As
in the case of the shallow wells, a pump test  will be required to
establish the  exact number, capacity, and location of the deep
extraction wells.

ON-SITE GROUNDWATER TREATMENT
  An on-site treatment facility will be used to reduce the level of
volatile organic contamination in the groundwater to be extracted
by the shallow wells to a level acceptable for discharge to the
Atlantic County Utilities Authority (ACUA) wastewater treatment
plant (WWTP) for further treatment. The  less contaminated
(deeper) groundwater will bypass the on-site treatment system, will
be combined with the treated groundwater and the total flow will
be pumped through a force main to  an existing sewer for  con-
veyance to the ACUA treatment plant for final treatment prior
to disposal.
  A process flow schematic of the  proposed on-site pretreatment
system is provided in Figure 17. The primary factors  affecting
design of the  on-site groundwater treatment facilities include the
following: (1) anticipated initial VOC concentration, (2) expected
decrease in contaminant concentrations with time, (3) the types
of contaminants to be pretreated and (4) the need to assure adequate
safety for operators and nearby residents. Upon consideration of
these factors,  the treatment train shown in Figure 17 was adopted.
The proposed on-site treatment system for the shallow groundwater
will consist of the following:

• pH reduction by addition of H2S04
• Air stripping
• Vapor phase carbon adsorption
• Liquid phase pumping to ACUA WWTP.

  The groundwater treatment facilities will be designed to  treat
an average of 200,000 gal/day of extracted  groundwater at an
average VOC removal efficiency of 96 to 98%, based on an initial
influent VOC concentration of 50 mg/1. The treated groundwater
will  contain an average of 1 to 2  mg/1 total VOC.

pH reduction
  Groundwater to be extracted from the shallow aquifer is expected
to contain substantial concentrations of dissolved iron. To prevent
iron precipitation and clogging of air  strippers, the pH of the
extracted groundwater would be reduced from about 6.5 to 7.5
to about 3.5 by the addition of sulfuric acid prior to air-stripping
pretreatment. At this pH, iron will remain in the dissolved state,
and iron precipitation and biological fouling of the packing medium
is not expected. Even though no biological fouling is expected at
the pH of 3.5,  provision would  be made  to clean the towers
periodically with a sodium hypochlorite solution.

Air  Stripping
   Based on the results of pilot studies, air stripping was selected
as the preferred process for removal of volatile organic compounds
                                                                              MODELING OF HAZ MAT TRANSPORT     117

-------
                                                                                            PROPOSED GROUHOWATER MONITORING WELLS

                                                                                    -.•'l| ฉ  SHALLOW EXTRACTION WELLS

                                                                                        0  DEEP EXTRACTION WELLS
                                                                Figure 16

                                           Proposed Location of Monitoring and Extraction Wells
                                                                    to >tMOsm*iit
                           OrSCMAIKC
                         TO ITuOVKOt
         PHปSฃ
       TRCATUCNT
Vii-OK
?MASC
CU40H


?
ซ-
v*POป
nusc
cซซstป
3 "ปY—'

 lltp.l
                                                               L7-
                                                                       r
           SHปLLOป
            •tLL
            PUUPS
G-
                            /-KSn
                                            BlOปtl>
                                             (TTP.I
            DCCP
            •Ell
            PUUPS
                                                          -•/
                                                          I     ^  .
                                                       /   \  AltiJHION
                                                                Figure 17
                                          Process Glow Schematic for On-Site Pre-Treatment System

                                                                                                   ursi
                                                                                                   CIMK4TOH

                                                                                                    '
                                                                       .H0221CS
                                                                        ItYPJ
                                                                                                      .
                                                                                                    (trp.1
                                                                                                             PUMP
                                                                                                                         T<>
118     MODELING OF HAZ MAT TRANSPORT

-------
from the groundwater to be extracted from the shallow aquifer
at Price's Landfill No. 1. Three parallel packed towers including
a standby unit will be provided, each capable of treating 50 per-
cent of the design  flow of 200,000 gal/day. The  towers will be
designed to achieve an average removal efficiency of 96 to 98%
based on an initial influent VOC concentration of 50 mg/1 and
an average effluent VOC concentration of 1 to 2 mg/1. Both towers
will be 3 ft. in diameter with a total height of 25 to 30  ft. and a
packing depth of 20  ft. The surface loading rate  on each tower
will be 10 gal/min/ft2 with an empty bed contact time of 7.5 min.
These preliminary sizes and flow rates would be finalized during
design of the pretreatment  facilities.
  Air will be supplied to the stripping towers by rotary forced-
draft blowers. Two blowers will be provided, each sized to provide
the design air flow rate at an initial average air-to-water ratio of
80-to-l. Under normal operation, one blower will be in service and
the other will  serve as a standby unit.  To obtain high removal
efficiency, spray nozzles will be used to atomize the groundwater
flow at the top of the packing towers. Tri-packs (2-in) will be used
as packing medium for the towers.

Dehumidification
  The solvent laden  air  to  be discharged from the air  stripping
towers will flow through an overhead condenser and electric coil
heater for dehumidification prior  to  carbon treatment.  The
dehumidification system will be designed to reduce the humidity
to about 40 to 50%  relative humidity at an air temperature of
approximately  100 axaF. Depending on the temperature of the
ambient air, either heating coils and/or overhead condensers will
be required for proper dehumidification on a seasonal basis.
  The condensed waters would be collected in 55-gallon drums
(predominantly in the summer) and stored on the premises.
Periodically, the contaminated waters will either be recycled during
cold temperature periods or transported  off-site to a RCRA-
permitted disposal facility.

Vapor Phase Carbon Treatment
  A number of process parameters were evaluated to determine
the optimum conditions for removal of VOCs in the discharge air
stream from the stripping towers. The parameters considered in-
clude (1) type of carbon to be used, (2) influent and effluent VOC
concentrations, (3) the competitive nature of the constituents with
respect to the carbon, (4) anticipated carbon usage rate, (5) bed
contact time and (6) the method of carbon regeneration.
  Initially, it is anticipated that the air stripping towers will remove
as much as 80 Ib/day of total volatile organic compounds from
the influent groundwater. The anticipated amounts of volatile
organic compounds  remaining  in the  air discharge from the
stripping towers normally will exceed the limitations set by NJDEP
for air stripping towers (0.10 Ib/hr based on 24-hr operation, or
2.4 Ib/day). In order to meet the mandated emission rates, a vapor
phase carbon adsorption system will be required to treat the air
stream exiting  the stripping towers  prior to discharge to the
atmosphere.
  Based on the anticipated initial concentration and characteristics
of the tower off-gas, it is proposed that two dual carbon beds for
each tower be provided for treatment. The rate of expenditure of
carbon may range from 500 to 1,000 Ib/day. These expenditure
rates are considered acceptable for off-site regeneration of carbon.
Effluent Pumping
  Three constant speed pumps with a capacity of 450 to  500
gal/min each will be provided to convey the extracted groundwater
from Price's Landfill No. 1  to the ACUA plant for further treat-
ment and disposal. For the first 5 years of operation, two pumps
will be used continuously with the third pump used as a standby
unit. After 5 years, when the two deep wells are anticipated to be
taken out of service, the reduced flow will be pumped by one 450
gal/min pump on an  intermittent  basis.

CONCLUSIONS
  Groundwater in the sand and clay layers of the upper Cohansey
formation was found to be heavily contaminated with organic and
inorganic chemicals disposed of at the Price's Landfill No. 1. Field
sampling, analysis and groundwater modeling of flow and con-
taminant transport indicated that remediation at Price's Landfill
No. 1 can be effectively accomplished by pumping, extraction and
treatment of the contaminated water. On-site treatment would
consist of pH adjustment, air stripping in packed towers and vapor
phase carbon adsorption. Treated effluent from the on-site plant
would be pumped to the ACUA wastewater plant for further treat-
ment and disposed to meet  State discharge requirements.

REFERENCES

1.  Camp Dresser & McKee Inc.ฎ "Remedial Investigation and Feasi-
   bility Study for Price's Landfill No. 1," prepared for New Jersey Depart-
   ment of Environmental Protection, February, 1985.
2.  Camp Dresser & McKee Inc. "Price's Landfill No. 1 - Conceptual Design
   Report," prepared for the New Jersey Department of Environmental
   Protection, March, 1987.
                                                                             MODELING OF HAZ MAT TRANSPORT     119

-------
                    Trend-Surface  Modeling  of Groundwater  Data
                                                Charles  T. Kufs, CPGS
                                                 Roy F. Weston, Inc.
                                                   West Chester, PA
 ABSTRACT
   Evaluating the occurrence and migration of groundwater con-
 tamination often requires a quantitative model in which the relative
 accuracy and precision of predicted contaminant concentrations
 can be determined. One empirical approach that has been used
 successfully in developing  such models is called trend-surface (or
 response-surface) analysis. Trend, surface modeling is a statistical
 method for developing an equation  for predicting a spatially-
 distributed variable (e.g., contaminant concentrations, hydraulic
 heads, aquifer properties) from geographic coordinates. Trend-
 surface models can be used for a variety of purposes including in-
 terpolating between actual data measurements, identifying data
 anomalies and establishing confidence intervals around predictions.
   This paper describes  what trend-surface modeling is, how to
 develop a trend-surface model and evaluate its accuracy and pre-
 cision and how trend-surface modeling can be used to provide in-
 novative solutions to  problems in groundwater monitoring.

 INTRODUCTION
   A common practice in hydrogeology is to draw contour lines
 to depict the occurrence of spatially-distributed (i.e., regionalized)
 variables such as groundwater elevations, aquifer properties (e.g.,
 transmissivity) or contaminant concentrations. Where the occur-
 rence of such variables must be evaluated quantitatively, the statisti-
 cal method known as trend-surface modeling (or response-surface
 modeling) can be used.
   Trend-surface modeling  is a simple extension of multiple regres-
 sion in which a spatially-distributed variable is represented by an
 equation consisting of  algebraic functions of sample location
 coordinates'. Trend-surface models  smooth data irregularities
 caused by natural variability or random errors to identify regional
 trends.
   Trend-surface modeling of groundwater data can serve a variety
 of purposes including:

 • Depicting regional  trends  in groundwater elevations, aquifer
   properties or contaminant concentrations
 • Predicting values of a regionalized  variable at locations where
   there are no actual measurements (e.g., site boundary, point of
   compliance)
 • Estimating possible locations of undocumented contaminant
   sources
 • Evaluating the adequacy of sample (e.g., monitor well) locations
 • Identifying possible anomalies  in groundwater or contaminant
   flow
 • Assessing the uncertainty in aquifer properties or contaminant
   concentrations to examine the relative influence of regional
  trends versus natural variability
• Calculating the range of contaminant concentrations likely to
  be observed at a point.

  Figure 1 illustrates isometric diagrams of several relatively simple
trend surface models for the migration of a groundwater con-
taminant.

CONVENTIONAL TREND-SURFACE MODELS
  Any straight line formed by plotting related y and x values can
be expressed algebraically by the equation:

  y = aQ + a,(x,)                                       (1)

where ao is  the point at which the line  crosses the y-axis and a,
is the slope of the line. Linear statistical relationships contain a
certain amount of scatter between points, called error (e), so that
statistical models are expressed by:
  y =  a0  + a,(x,) + e
                                                     (2)
Statistical relationships can be extended to many dimensions (i.e.,
"x" axes), as expressed by the model:
  y =  a0+ a,(x,) +
                                                     (3)
  This multidimensional model also can be expressed by the short-
hand notation:
y =  a0
            f(x, x2   • •  xn)
(4)
  which indicates that y is related to the sum of a constant (aj,
some mathematical function of "n" number of x variables and
random error.
  Trend-surface models relate regionalized (i.e., spatially, depen-
dent variables, such as contaminant concentrations or groundwater
levels, to the  locations at  which the variables are measured such
that:
            f(location)
                                                     (5)
  The location function is expressed by geographic coordinates
(e.g., northing and easting) typically but not necessarily in the form:
  f(location)  = a, (N)  + a, (E) +  a3 (NE) + e
                                                     (6)
  This trend-surface model is a first-order model; the order of the
model is equal to the highest power of the location coordinates.
Thus, a second-order model would be written:
  f(location)  = a, (NJ  + a2 (E2) + a3 (N) +  a4 (E)
               +  a, (NE)2 + a6 (NE) +  e
                                                     (7)
120    MODELING OF HAZ MAT TRANSPORT

-------
                                                                                                             Shallow Zone
                     October 1982
                                                          October 1983
                                                                                             February 1985
                                                             Figure 1
                                        Isometric Plots of Trend Surfaces for Contaminants in the
                                          Shallow, Intermediate and Deep Zones of an Aquifer
  Typically, conventional trend-surface models seldom have an
order higher than four. Commercially available computer programs
for generating trend-surface models usually are limited to second-
order models.
  Another limitation of trend-surface models is the number of data
points required. As a rule-of-thumb, at least 10 values of the region-
alized variable are needed for each location term in the model to
calculate reliable values for the model coefficients (i.e., the "a"
values in the equations). Thus, a first-order model might require
30 data points while a second-order model might require 60 points.
Reliable trend surfaces can be generated from fewer than 15 points,
however,  if the random error is small.

OPTIMIZED TREND-SURFACE  MODELS
  Conventional trend-surface models have been found to be satis-
factory for most applications. In some  situations,  however, con-
ventional  trend-surface  models  tend to be inefficient (i.e., some
of the location terms in  the model do not significantly reduce the
error). In these cases, additional steps  can be taken to optimize
the fit of  the model to the measured data. An approach recom-
mended to develop an optimal trend-surface model is to:

  Verify the data
  Evaluate the frequency distributions
  Select a location-coordinate system
  Select model components
  Generate  the trend-surface model
  Evaluate the degree-of-fit
  Apply the trend-surface model

Step 1—Verify the data
  The first step in the process is to verify that the computerized
data are correct. Anyone who has  had to re-analyze a data set
because of a single error will appreciate the value of this step. Con-
sider  this—given   four-digit coordinates  and  a  three-digit
regionalized variable, a  small (e.g.,  15 observations) data set will
have over 100 numerical entries such that a 99%  accurate data-
entry specialist probably will make two errors.
  Data should be verified at the points of generation (i.e., was the
laboratory, surveyor or  field technician correct?), archiving (i.e.,
was the data transcribed  and computerized correctly?) and manipu-
lation (i.e., did the programming function as planned?). Checks
of both individual data points and summary statistics or plots
should be completed.

Step 2—Evaluate Frequency Distributions
  One of the underlying assumptions in trend-surface modeling
is that the measurements of the regionalized variable (i.e., the "y"
in the previous equations) approximate a normal (i.e., bell-shaped)
frequency distribution. This assumption usually is correct in the
case of groundwater elevations but not in the cases of aquifer
properties or contaminant concentrations. Where the assumption
is not met, confidence limits and statistical tests of the model will
be biased. The validity of the  assumption can be evaluated using
combinations of descriptive statistics (e.g., skewness,  kurtosis),
graphics (e.g., histograms, stem-leaf plots, box plots, probability
plots) and goodness-of-fit tests. If the data are positively skewed,
as is common with contaminant concentrations, a logarithmic trans-
formation of the data  generally is  beneficial.
  Figure 2 illustrates the improvement in four histograms after a
common-log transformation.  Unfortunately, because logarithms
are nonlinear, transforming back to the original scale of measure-
ment is not  always straightforward.

Step 3—Select a Location-Coordinate System
  Existing geographic-coordinate systems seldom coincide with the
axes of minimum variation in a  regionalized variable. Conse-
quently, to optimize the efficiency of a trend-surface model, a trans-
formed coordinate system should be identified that will minimize
the need for additional location variables in the model. Trans-
forming locations on an existing coordinate system (N, E) to a new
coordinate system (N', E') involves an angular rotation (ox) and
linear translations (tn,  te) according to the formulas:
N,  = (N  - tn) cosine -  (E  - te) sine

E,  = (E -  tj cosine + (N - tn) sine
(8)

(9)
  Selecting the best coordinate system is a  matter of visually
judging which of several rotations and translations of the original
coordinates will minimize the scatter in the regionalized variable
or reposition the origin of the coordinate system at a mathemati-
cally convenient point. Figure 3 illustrates the selection of a trans-
                                                                              MODELING OF HAZ MAT TRANSPORT    121

-------
     Before Log Transformation*
After Log Itimlomutkm*
                                                                    Rotallon Angle
                                                                                        Form ol Data Plot
                      = 29.8
                                                          S = *0.3
                                                          K = 1.1
                    S= -37
                    K= 206
                                                          S- -1.0
                    S= -25
                                                          S--09
                                                          K =06
                    S= -24
                    K = 4.5
1 = Normal Distribution: Skewneu = 0.0; Kurtoปiป = 3.0

                            Figure 2
             Examples of Log Transformations to Correct
                      Poorly Distributed Data
                                                          K -0.1
                                                                   No Rolalion   y
                                                                   25 Degrees
                                                                   45 Degrees
                                                                   65 Degrees
                                                                   90 Degrees  y
                                                                                       25
                                                                                             50
                                                                                                  75   100
                                                                                                       100
                                                                     Best Transformation
                                                                     notation 45 Degrees
                                                                     Translaiion -50 Feel
                                                                     Curve Form y  -(ซ') Where
                                                                              a is > 1 and ev
                                                                                       25
                                                                                             SO
                                                                                                  75
                                                                                                       100
                                                                                       25
                                                                                             50
                                                                                                  75
                                                                                                       100
                                        I    25    50   75   100

                                          Location Coordinale (Feet)

                                                      Figure 3
                                         Transformation of Location Coordinate
                                               Based on Data Symmetry
 formation based on data scatter and symmetry. Figure 4 illustrates
 the selection of a transformation based on the functional form (in
 this case, probably cubic) of the data.

 Step  4—Select the Model Components
   Surfaces  representing the occurrence of groundwater or con-
 taminants maybe smooth but are seldom flat. Consequently, these
 surfaces must be represented by nonlinear mathematical functions.
 Conventional trend, surface models use polynomials (i.e.,  x2, x3,
 x4); however other nonlinear functions also can be used.
   The simplest method of selecting possible model components is
 to plot the regionalized variable (e.g., contaminant concentration)
 versus the location coordinates (i.e., N, E and their product, called
 an "interaction," NE) and  then visually selecting the mathemati-
 cal functions that will best linearize the data plots. Figure 5 illus-
 trates several transformations  for  correcting  nonlinear data
 relationships. Correlation coefficients can be used in conjunction
 with the plots to select candidate location functions for the model.

 Step  5—Generate the Trend-Surface Model
   Given the candidate location functions selected for the model
 in Step 4, the next step is to develop a model that uses only those
 functions that significantly add to the prediction of the regional-
 ized  variable.  Conventional trend-surface models utilize  all the
 terms of the polynomial expansion, whether the terms add to the
 model or not, so  that the models are sometimes  relatively ineffi-
                          cient.  Trend-surface models  can  be improved,  therefore,  by
                          including only those location functions that are highly correlated
                          with the regionalized variable but are themselves  uncorrelated.
                            Probably the best approach to generating the trend-surface model
                          is to utilize a stepwise regression algorithm in which location func-
                          tions are sequentially added to (or deleted from) the model until
                          all the candidate functions have been evaluated. The best model
                          then is selected  based on some statistical criterion, such as the
                          proportion of variance accounted for (i.e., R2). Table 1 summa-
                          rizes the statistics most commonly used to evaluate models.
                          Step 6—Evaluate the Degree-of-Fit
                            As with all models, it is important to establish the  extent to which
                          a trend-surface model  accurately  represents  the  regionalized
                          variable, that is, the degree-of-fit2'3.  Model evaluation generally
                          involves examining:

                          • Overall Model—Does the model (i.e., the equation) account for
                            a large proportion of  the variation in a regionalized variable,
                            with a minimum of uncertainty
                          • Model Components—Does each location function contribute sig-
                            nificantly to the model?
                          • Individual Data Points—Do any data points exert an unaccept-
                            ably large influence on the model (i.e., are there any outliers)?

                            The overall model can be evaluated using  statistics and plots
                          summarized in Table 1.  Model components can be evaluated us-
 122    MODELING OF HAZ MAT TRANSPORT

-------
Rotation Angle
 No Rotation y
                  Torm of Data Plot
 10 Degrees y
 20 Degrees y
 30 Degrees y
 45 Degrees y
Best Translormation
Rotation: 45 Degrees
Translation: -125 Feet
Curve Form: y  -x1 Where
         a is > 2 and odd
                     100   150
                                200
               Location Coordinate (Feet)
                            Figure 4
               Transformation of Location Coordinate
                     Based on Form of Data
                                                                        Form ol Data Plot

                                                                                Large x
                                                                         Large x"-	
                                                                                               Transformation
                                                                                                   x"
                                                                                                 10-; e-
                                                                                                  1/x-1
                                                                                                               Numerical Conditions*
                                                                    Pos
                                                                    Pos
                                                                    Neg
              Pos

            Pos; Even
                                                                                                 Log x; Lnx
                                                                                                 a/7"
                                                                                                   x"
                                                                                                  -M
                                                                    Pos
                                                                    Pos
                                                                    Neg
                                                                    Neg
              Pos
            Pos; Odd
            Pos; Even
                                                                                                  -M
                                                                                                  -(10-)
                                                                    Pos
                                                                    Pos
                                                                    Pos
                                                                    Neg
                                                                                                                             Pos
                                                                                                                           Pos; Even
                                                                                                  1/x"
                                                                                                -Log x; -Lnx
                                                                                                   x"
                                                                                                  -(x')
Pos
Pos
Pos
Neg
Neg
  Pos

  Pos
Pos; Even
Pos; Odd
                         •Pos: Greater Than or Equal to 1
                          Neg: Less Than or Equal to -1

                                                    Figure 5
                            Selected Transformations for Correcting Nonlinear Relationships
ing statistics and plots summarized in Table 2. Individual data
points can be evaluated using statistics summarized in Table 3.
  To illustrate how the previous optimization steps can produce
a more accurate, precise and efficient trend-surface than conven-
tional models, Table 4 summarizes comparative statistics for models
generated for chloride concentrations in the groundwater near a
landfill4. The optimized trend-surface  model accounts  for the
most variation (86%) with a lower ratio of error to mean value
(18.6%) than any of the conventional models while using the same
number of terms as a conventional third-order model.

Step 7—Apply the Trend-Surface Model
  Once generated, a trend-surface model can be used for a variety
of purposes. Most commonly, trend surfaces are used to depict
graphically the patterns formed by the distribution of a regionalized
variable. However, a better use for such a quantitative model is
to predict values of a regionalized variable at locations that have
not  been sampled.  For  example, trend-surface predictions of
groundwater elevations, aquifer properties or contaminant con-
centrations  can  be used  as  input  to groundwater  flow or
contaminant-transport models that require information on these
parameters at numerous points on regular grids. The uncertainty
in these predictions can also be calculated to provide insight on
the reliability of the predicted values.
                         INNOVATIVE TREND-SURFACE MODELS
                           Although the underlying strategy of trend-surface modeling is
                         to relate a regionalized variable to a mathematical function of
                         location coordinates, there is no requirement that the model con-
                         sist only of location functions (i.e., functions of N and E). There-
                         fore, it may be appropriate in some situations to add terms to the
                         model, called covariates, that are themselves spatially-distributed
                         variables. For example, the concentration of a contaminant  in
                         groundwater could be predicted from a model consisting of both
                         location coordinates and groundwater elevations, soil-gas measure-
                         ments   or   geophysical  (e.g.,  electromagnetic  conductivity)
                         measurements5. Table 5 summarizes examples of covariates that
                         could be useful in some trend-surface models.
                           Because groundwater elevations and contaminant concentrations
                         vary over time as well as location, a trend-surface  model can be
                         fine-tuned at selected locations by using another type of statistical
                         model, called  a time-series model, to eliminate temporal  trends
                         from the spatial error. Combining trend-surface and  time-series
                         analyses is called statistical dimension modeling.
                           Statistical dimension models for contaminant concentrations (C)
                         can be expressed by the equation:

                           C =  ac + f(location) +  f(seasonality) +  f(time trend) +  ec
                                                                                  (10)

                                    MODELING OF HAZ MAT TRANSPORT     123

-------
                             Table 1
              Selected Statistics for Evaluating Overall
                       Trend-Surface Models
  Coefficient of
  Determination (H )
  F-test and
  Probability
  Standard Error
  of Estimate
  (Root Mean)
  Square Error)
  Mallow's Cp
  Criterion
  Plot of Actual
  Versus Predicted
  Values
                          The maximum proportion of the total
                          variance  in  y Jihat the  model   can
                          account for.   R  can be adjusted to
                          compensate for the  number of termj
                          in the  model and sample  size.    R
                          should   be   at   least   0.6    and
                          preferably between  O.B  and 1.0  for
                          a reliable trend-surface model.

                          Essentially a test of whether
                          R  is statistically greater than 0.
                          The F-value  will vary  with sample
                          size and the number of terms in  the
                          model but the probability should be
                          less than  0.1  and preferably below
                          0.05.

                          The standard deviation of the
                          residuals  (i.e.,   the  difference
                          between the  observed  and predicted
                          values).  The standard error of
                          estimate should  be  less  than about
                          30%  of   the  acceptable  range   for
                          values  predicted from the  trend-
                          surface model.

                          A relative measure  of the  bias  in
                          the model.  Ideally, Cp should be
                          small and  close  to the  number  of
                          terms in the  model.

                          Ideally, points  should plot  close
                          to a 45-degree line passing through
                          the origin.  Systematic deviations
                          from the  line  indicate  a  lacfc-of-
                          fit  with  the  model.   Individual
                          points deviating  from  the  line   may
                          be considered outliers.
                                                                        Residuals
                                                                        Leverage or
                                                                        Hat Diagonal
                                                                                                 Table 3.
                                                                                    Selected Statistics for Evaluating Data
                                                                                   Used to Develop a Trend-Surface Model
                                                                                           The difference  between the observed and
                                                                                           the predicted  value*.  Residuals  should
                                                                                           be small and uniformly distributed.

                                                                                           A measure  of whether an  individual "x"
                                                                                           value  may be an outlier  from  the model.
                                                                                           The   leverage    for   an    individual
                                                                                           observation  should   be   less  than  two
                                                                                           times  the  number of  terms In  the model
                                                                                           divided  by the sample size.
                                                                      Studentized
                                                                      Deleted Residual
                                                                        Cook's Distance
                                                                        Measure (C.)
                      A measure  of whether an  individual  "y"
                      value may be  an  outlier from the model.
                      The studentized  deleted residual is  like
                      a t-statistlc that  will vary with sample
                      size.

                      A measure  of the overall  impact  of  an
                      Individual observation on the
                      coefficients  of  the  trend-surface model.
                      If C.  is less than  0.2,  the observation
                      has little influence on the modelr if  C.
                      is greater than  0.5,  the  observation  ii
                      very influential.

                      The changes In the  regression
                      coefficients  that   would   result   from
                      deleting  an   observation.  The  DFBETA'e
                      should  all   be   small  and   relatively
                      consistent across observations.
                                                                                                 Table 4
                                                                          Comparison of Conventional and Optimized Trend-Surface
                                                                        Models for Predicting Chloride Concentrations in Groundwatcr
 Plot of Actual
 Values Versus
 Residuals
                          Ideally,  points  should  plot as  a
                          horizontal band  of  constant thick-
                          ness.   Non-horizontal trends
                          generally  indicate  the  need  for
                          additional  terms   in   the  model.
                          Non-uniform  thickness  trends  may
                          indicate  that  the variance  is not
                          constant, thus  requiring  a  trans-
                          formation of the y values.
                            Table 2
           Selected Statistics for Evaluating the Elements
                    Of a Trend-Surface Model
Regression
Coefficients
t-test  and
Probability
Standardized
Regression
Coefficients
Variance
Inflation
Factor
Partial
a Regression
Leverage
Plots
                     The coefficients (i.e.,
                     associated with the "x"
                     trend surface model.
                                               the "a" terms
                                               term*)  in the
                      A   test   of   whether    a    regression
                      coefficient is statistically
                      different  from 0.  The t-value  will  vary
                      with  sample  size  but  the   probability
                      should be  less than 0.1

                      The  regression  coefficient  divided  by
                      its    standard    error.     Standardized
                      regression coefficients can  be used  to
                      evaluate   the  relative   importance   of
                      Individual predictor  terms In the model.

                      A  measure  of   how  much   the  model's
                      coefficients     change    because     of
                      correlations   between    the   predictor
                      variables. The variance inflation factor
                      should be less  than   10  and  preferably
                      near 1.

                      Plot of the regionalized  variable versus
                      predictor  variable with  the effects  of
                      the  other  predictor   variables  in   the
                      model  held constant.  The slope  of  the
                      line fit to these  data is the regression
                      coefficient    for    that     predictor
                      variable. These plots provide Insight on
                      the   variability   of  the    regression
                      coefficients.
                                                                     CQV AR I ATI

                                                                     Saeplt Dซpth
                                                                                                  Table 5
                                                                                Examples of Covariales That Ma) Be Useful in
                                                                                            Trend-Surface Models
                                                                                                                       REPEATED
                                                                                         POSSIBLE DSE IN PREDICTION QP     MEASUREMENTS
                                                                                       Hater   Aquifer     Ground Hater   REWIRED FOR
                                                                                                          r"1t^Tif M* ion  PQReCASTTMC*
                   7

                   Tee
                                                                                                  To

                                                                                                  Ho
                                                                                                            Tee
                                                                                                                          no

                                                                                                                          Ho
Surface Blevatloni
Topography

Climatic Paraปetera  Yea
(e.g., precipitation,
teeperature)

Ground Mater         —
Level!

Geotechnlcal         Yea
Paraatet en (e.g . ,
denalty, poroaity,
•end/clay content  of
aquifer Mterlala)

Ceophye!cซl          Tei
HeaaurcBente  (e.g.,
electromagnet Ic
conductivity)
                                                                     Soll-gae             7        No         Yea            Y.,
                                                                     Neaaureeents

                                                                     Contulnet Ion        No        Ho         Yet            yea
                                                                     Surrogates  (e.g.*
                                                                     pa, apeclflc
                                                                     conductance,
                                                                     teapcrature)

                                                                     Tlaปe Parameters      Tee        Ho         Yea            yta
                                                                     (e.g.,daya  froe
                                                                     arbitrary starting
                                                                     point, (eaional
                                                                     rector)

                                                                     • - Rapeated ••••urenentii required for t i me -dependent covarlatee
                                                                     b- Inorganic ContanlnantB
                                                                     c- Volatile Organic Contปlnanti
124     MODELING OF HAZ MAT TRANSPORT

-------
  In this equation, ac is the average of all the contaminant con-
centrations, f(location) is the trend-surface model, ec is the error
remaining after spatial and temporal trends have been removed
and the two time-series terms represent mathematical formulas to
account for the season the sample was collected and the long-term
trend in the contaminant concentration.
  The long-term change in contaminant  concentration is repre-
sented by the time-trend function which is simply the number of
days (tt)  from an arbitrary starting date (t0), expressed by:

  f(time trend) = a, (t0  + t,) + e,                       (11)

  The seasonally function is expressed by a trigonometric function
of an adjusted time function. The time function consists of the
initial time (t0) adjusted by adding some  number of days (ta) to
bring the trigonometric function in phase with the seasonal changes
in contaminant concentrations. Thus, the seasonally function can
be expressed by:

  f(seasonality)  = as[sin(t0 + tj  +  cos  (t0 +  ta) ] + es   (12)

  By using a statistical dimension  model, it is possible to deter-
mine whether differences in contaminant concentrations can be
attributed to location, seasonal fluctuations, long-term time trends
or random error4.

CONCLUSIONS
  Trend-surface models can be used to quantify the spatial distri-
bution and uncertainty of regionalized variables such as ground-
water  elevations,  aquifer  parameters  and  contaminant
concentrations. Conventional trend-surface models use polynomial
functions of location coordinates; however, better results often can
be obtained through  the use  of other nonlinear mathematical
functions. Important  steps in  generating a trend-surface model
include the verification of the raw data, the selection of the loca-
tion, coordinate system and the functional components of the
model and the evaluation and use of the model. Innovative trend-
surface models can be developed that include covariates such as
geophysical   measurements.  An  extension of  trend-surface
modeling, called statistical dimension modeling, can be particu-
larly useful for parameters that vary over both location and time,
such as groundwater elevations and contaminant concentrations.
REFERENCES
1.  Watson, O. S. Trend Surface Analysis. /. Math. Geo. 14,1971,161-187.
2.  Belsley, D. A., Kuh, E. and Welsch, R. E. Regression Diagnostics. John
   Wiley and Sons, New York, NY, 1980.
3.  Neter, J., Wasserman, W. and Kutner, M. H. Applied Linear Regres-
   sion Models. Richard D. Irwin, Inc., Homewood, IL, 1983.
4.  Fletcher, R.  L., Jones, A.,  Sheedy, K. A., Kufs, C. T. and Stein,
   D. M. "Quantitative Evaluation of a Groundwater Monitoring System,"
   Proc. of the Tenth Annual Madison Waste Conference. University of
   Wisconsin-Madison, Madison, WI,  1987.
5.  Folsom, R. E.  "Sampling and Modeling Superfund Site Pollutant
   Plumes: Methods Combining Direct Measurements and Remote Sensing
   Data," Proc. of the ASA /U.S. EPA Conference on Sampling and Site
   Selection in Environmental Studies. Washington, D.C.,  1987.
                                                                              MODELING OF HAZ MAT TRANSPORT     125

-------
                          The  Impacts  of  Using Assumed  Versus
                                            Site-Specific Values
                              In  Determining  Fate  and  Transport

                                               Pinaki  Banerjee,  Ph.D.
                                              David H. Homer, Ph.D.
                                      PRC Environmental Management, Inc.
                                                   Chicago, Illinois
ABSTRACT
  The quality of risk assessment studies is contingent upon the
availability of site-specific parameters used to estimate exposure
potential.  At Superfund sites, parameters such as total  organic
carbon, Eh, pH, and dissolved oxygen, often are not available,
especially for subsurface environments. In these instances, scientists
routinely incorporate assumed values into quantitative or semi-
quantitative estimation techniques. Also, predictions of fate and
transport of contaminants are often discussed in very generalized
terms for all possible reaction pathways.
  This paper emphasizes the importance of obtaining measured
or observed site-specific parameters and compares the use  of these
values to the use of assumed values. It recommends that remedial
investigation reports include provisions  for obtaining these
important parameters.

INTRODUCTION
  The SARA requires that risk assessments be conducted  at every
Superfund site  after RIs are conducted so that remedial actions
are optimized. A major objective of assessing the risks of releases
from  a  Superfund site is to determine the  actual  or potential
endangerment to human health and  the environment.
  The central focus of an endangerment assessment is the extent
of present or future exposure to  contaminants released from the
site. Assessing risks is an easy task when exposure concentrations
at receptor locations have been sampled and contaminants con-
centrations verified; one has actual data on which to base the risk
characterization. However, the risk assessment team often  is asked
to predict  future exposures in order  to protect future receptors.
These exposure estimates are often the most  disputed portion of
a risk assessment.
  Exposure estimates often are produced by mathematical  models,
both simple and complex, that use measured values, assumed values
or literature values as input. The reliability of outputs from these
models depends on the quality and quantity of the available data.
Site-specific information  that  help in  quantitatively, semi-
quantitatively or even qualitatively evaluate exposure potential is
not always available. In these instances, it is more difficult for the
risk assessment team to accurately assess site releases.
  This paper focuses on the data needed to estimate,  as accurately
as possible, the future potential exposure to contaminant releases
from a Superfund site. To estimate potential exposure dosages,
one must evaluate the fate and transport processes that determine
a pollutant's probable pathways  to,  and environmental  concen-
trations at, points of exposure. These fate and transport processes
include  sorption, complexation, volatilization, bioa
-------
Jersey coastal plane aquifiers, TOC values were reported at 1.3%
and 2.6%".
   Depending on the TOC value assumed, the estimated sorption
potential can greatly vary.  Furthermore, it has been well docu-
mented in the literature that sorption estimates based on TOC
content are not applicable to most aquifers because the TOC
content is less than 0.1%. Care must be taken in using models that
base sorption estimates on carbon content for such aquifers.
   To illustrate the impact TOC values  may have  on transport
estimates, we examine its influence on one coefficient often used
to estimate attenuation  of contaminants in groundwater—the
retardation factor, (R), that measures the retardation of a pollu-
tant relative to water. The retardation factor may be expressed as:

            R   =   1+dk                              (1)
                        v
   where
            d   =  soil  bulk density
            v   =  volumetric water content
            k   =  sorption coefficient

   The parameters d and v exhibit a narrow range of values; there-
fore k, or the sorption  coefficient, has the greatest effect in
estimating R. This is illustrated in Table 1, which calculates R using
different TOC values and assuming representative values for d and
v as 1.6 and 0.4, respectively. The compound chosen for these R
calculations is xylene (mixed), which has a KOC value of 24013.
Table  1 % that when assumed TOC values of 5 and 1 % are used
(as is  often the case), R values are  calculated at 49 or 10.6,
respectively. These R values indicate that the pollutant would be
expected to move by factors of 49 and 10.6 slower when compared
to water. However, if one assumes a TOC content of 0.1% or
lower, which is more representative of most aquifers, the R value
is 1.96 or lower. Compounds with a retardation value of 2  or less
can be considered to move with the water in light of the  uncer-
tainties associated with KOC values. Therefore, if one assumes a
TOC content of 1 or 5% for the aquifer,  the contaminant is
expected to move at a rate slower than water by a factor of 10 to
50, while if, in reality,  the TOC content  is 0.1% or less, the con-
taminant is probably moving with the water. As stated earlier, if
the TOC content is less than 0.1 %, R values should not be estimated
from KOC.
                           Table 1
           Estimation of R Using Different TOC Values
     TOC
   (percent)
  K
(L/kg)
     5

     1

     0.1

     0.02
12

 2.4

 0.24

 0.048
49

10.6

 1.96

 1.19
  However, in media with a TOC content of less than 0.1 %, other
parameters such as total or swelling  clay content  influence
sorption5'6. For such media, the measured TOC content should
be used in the calculations, thus providing a conservative estimate
of a compound's sorption potential.
  Another important use of TOC values that should be mentioned
is in designing actions. Once again  it  is  important to know the
site'specific TOC values. If TOC content is underestimated, design
of remedial action based on pump and treat methods may under-
estimate the total load of contaminants to be removed.  For
example, forced flushing of contaminated soils in unsaturated zones
is a remedial option proposed or being used at many sites.  The
most important factor controlling mass removal of contaminants
from subsurface soils is the sorption potential of the contaminant
to the subsurface materials. Any estimate of flushing time  and
volume is significantly affected by the accuracy of the sorption
coefficient. If the TOC content in the unsaturated zone is expected
to be greater than 0.1%, a measured TOC value might provide
a better estimate of the sorption potential. If the TOC content is
less than 0.1%, it would be appropriate to assume a value of 0.1 %
so that a conservative estimate could be made.
   In addition to its importance in delineating extent of contami-
nation in surface or subsurface soils,  TOC content is also important
in estimating loss of a chemical due to volatilization from soils.
A method presented by the U.S. EPA14 for estimating the volatili-
zation rate of polychlorinated biphenyls from soils includes sorption
coefficients in the estimating procedure. A more accurate estimate
can be calculated via this procedure if the TOC value is site-specific.
Redox Potential
  Redox potential (Eh) is another parameter that can be routinely
obtained and that aids in understanding a compound's fate  and
transport. Site-specific redox potential data can more accurately
predict fate of heavy metals. The influence of redox potential on
the solubility and mobility of heavy metals has been well docu-
mented.  Reddy and Patrick9 observed a decrease in the concen-
tration of copper remaining in solution as the redox potential of
a sediment suspension decreased. In another study, Khalid, et al.7
indicted that the redox potential, along with pH, heavily influences
the transformations of cadmium. Under reduced and neutral pH
conditions, cadmium remains strongly sorbed to sediments but goes
into solution as sediments suspensions are oxidized.
  The influence of redox conditions on the potential for biotrans-
formation of organic contaminants is the subject of extensive on-
going research. A large number of compounds such as benzene,
toluene and xylene degrade under aerobic conditions15. In con-
trast, other compounds such as trichloroethene and tetrachlorethene
undergo sequential dehalogenation reactions under anaerobic con-
ditions. In the absence of site-specific data, any discussion of bio-
degradation is limited to whether a compound can potentially be
biodegraded or biotransformed under all conditions.  Hopefully,
current research on determining biodegradation rates will improve
the evaluation of exposure material.
  If Eh values are not available, dissolved oxygen concentrations
may be used to indicate the redox potential of groundwater.  For
example, if no dissolved oxygen is present, anaerobic conditions
and a reducing environment are probably present.

pH
  The influence of pH on the speciation of metals in natural waters
has been studied extensively. Together with Eh, pH controls the
solubility and, hence, the mobility of heavy metals in water  and
soil-water environments, the pH of natural waters determines the
presence or absence of certain inorganic ligands (such as sulfide
as opposed to sulfate),  which in  turn  affects complexation
mechanisms. Griffin and Shimp4 observed that sorption of heavy
metal cations such as lead, cadmium, mercury and chromium in
landfill leachate  increased at ph values between 5 and 6 due to
formation of insoluble hydroxide and carbonate compounds. They
concluded that heavy metals anions such as chromium (VI), arsenic
and selenium are significantly more mobile than cations of higher
pH.
  Thus,  the  pH of the environmental media is an important
parameter for predicting the fate of heavy metals and should be
measured routinely at hazardous waste sites.

Clay Content and Cation Exchange Capacity
  Clay content and cation exchange capacity also affect the  fate
of contaminants. Most chemicals preferentially sorb onto finer par-
                                                                            MODELING OF HAZ MAT TRANSPORT     127

-------
 deles rather than coarser particles. In addition, certain clay minerals
 such as montmiorillonite more effectively attenuate the movement
 of most chemicals as compared to  other clay minerals such as
 kaolinite or illite. Furthermore, as discussed earlier, sorption esti-
 mates based on TOC content are not valid for non-polar organics
 in most groundwater environments. Research indicates that for such
 groundwater environments, the surface area, the total clay content
 or the total swelling clay content may provide an estimate of the
 extent of sorption5'6. Particle  size  is one of the rare  physical-
 chemical parameters measured at most hazardous waste sites.
 However, surface area determinations and clay mineralogy analysis
 are complex and costly analytical processes. Furthermore,  analytical
 techniques to characterize clay mineralogy in amorphous subsurface
 zone material  have  not yet been developed.
   Analytical tools that help to characterize subsurface environ-
 ments are readily available. One such tool is a soil's cation exchange
 capacity (CEC). The CEC of soils may indicate the presence of
 specific clay minerals. Griffin and Shimp4 stated that CEC is the
 most significant property of a clay mineral affecting attenuation
 of heavy metals. The CEC, together with the clay content also may
 help evaluate the extent of sorption for organic compounds in soils
 with a low TOC content.

 Other Site-Specific Information
   In addition to the parameter  needs discussed  above, other
 information, such as climatological data, is required for some of
 the models. While developing exposure estimates, the risk assess-
 ment team often obtains information on climate from the nearest
 air station, which in  some instances can be a considerable distance
 from the site.  However, if prevailing wind directions are recorded
 during ongoing site activities, more accurate predictions can be
 made regarding the exposure potential for populations living off-
 site. Also, information on recreational or other uses of the site and
 surrounding areas, as well as physical descriptions of the site, is
 helps identify potential receptors and characterize the extent and
 rate of exposure.
   Information on other site-specific features also may be  necessary
 to estimate exposure. For example, if  surface water bodies are
 present at the site, estimates of the depth, area and water velocity
 are essential for estimating volatilization potential. Similarly, for
 sites containing waste piles or landfills, an estimate of the exposed
 surface area as opposed to covered areas is very useful.  For these
 parameters, field measurement  or even crude measurements are
 superior to  estimates based on visual observations.

CONCLUSION
  The technical adequacy of a qualitative or quantitative risk
assessment and, as a result,  the design of remedial actions, depends
on the availability of relevant site-specific information. To develop
the information needed for these tasks, researchers often  use
assumed values because site-specific values are not available.
However, the margin of error increases when assumed values are
used.
  The accuracy of most predictions is only as good as the  data
used to make predictions. Therefore, additional emphasis must be
put on obtaining site-specific information. At most sites,  such
information could have been obtained as part of ongoing activities
at little or no extra cost or effort if information  needs  had been
included in the original work plan. For example, soil samples are
routinely collected at Superfund sites for particle  size  analysis.
These same  samples  can be used to determine the TOC, Eh, pH,
and CEC content of soil. For groundwater samples, Eh, pH and
dissolved oxygen should be measured following well development.
Moreover, equipment  similar to that  used to measure  these
parameters in surface waters has been adapted for use in ground-
waters and is currently available.
  Obtaining  site-specific information will improve  estimation of
the extent of contamination and,  consequently,  improve the
effectiveness of remedial designs. Several research projects are being
conducted to adapt much of the physical-chemical knowledge
developed from surface water research and to apply that knowledge
to groundwater systems. The ultimate goal of these research efforts
is to  better  understand and  model  a constituent's subsurface
behavior. Precise modelling techniques may  not  be available,
however, for  several years. Moreover, when available, the accuracy
of these models will depend on the accuracy of the data now being
collected. As a result, remedial investigation work plans should
collect the site.specific  information  discussed in this paper.

REFERENCES
 1.  Banerjee, P.. M.D. Piwoni and K. Ebird. "Sorption of Organic Con-
    taminants to a Low Carbon Subsurface Core;"  Chemosphere 14:
    1985,1057-1067.
 2.  Banerjee, P., M.D. Piwoni and K Ebeid, "Sorption of Organic Solvents
    to Subsurface Soil, Presented at the Annual SETAC Symposium.
    St. Louis, MO, 1985.
 3.  Curtis, G.P. and Roberts, P.W., "Sorption of Halvgeneraled Organ-
    ic Solutes by Aquifer Materials: Comparison Between Batch Equilibri-
    um Measurements and Field Observations," Tech.  Report 285, Dept.
    of Civil Engineering,  Stanford University, Stanford, CA,  1985.
 4.  Griffin. R.A. and N.F. Shimp, " Attenuation of Pollutants in Municipal
    Landfill Leachate by Clay  Minerals." Municipal  Environmental
    Research Laboratory,  U.S. EPA, Cincinnati, OH EPA-600/2/78-I57,
    1978.
 5.  Hassett, J.J. et al. "Sorption of 2-naphthol: Implication concerning
    the Limits of Hydrophobic Sorption." SoiYSci. Soc. Am.}. 45:1981,
    38-42.
 6.  Karickhoff, S.W., "Organic Pollutant Sorption in Aquatic Sediments."
    J. Hyd. Engi.. 110 1984, 707-735
 7.  Khalid, R.A..  R.P. Gambrell.  and W,H.  Patrick. Jr.. "Chemical
    Availability of Cadmium in Mississippi River Sediment." J. Environ.
    Qual. 10 1981, 523-528.
 8.  Newson, J.M., Transport of Organic Compounds Dissolved in Ground-
    water. Groundwater Monitoring Review, Spring 1985, 28-36
 9.  Reddy, C.N. and W.H Patrick, Jr.. "Effect of Redox Potential on
    the Stability of Zinc and Copper Chelates in Flooded Soils." SoUSd.
    Soc. Am.  J. 1977 41  724-732
10.  Scwarzenbach, R.P. and J. Westall,  "Transport of Nonpolar Organ-
    ic Compounds from  Surface Water to Groundwater. Laboratory
    Sorption Studies."  Environ.  Geo. Tethurst.  15: 1981, 1360-1367
11.  Urchin, C.G. and G. Mangels. "Chloroform Sorption to New Jersey
    Coastal Plain Groundwater Aquifier Solids." Environ. Toxkol. and
    Chem. 5 1980. 5.339-343
12.  U.S. Department of Agriculture. "Soil Survey: Erie County, Penn-
    sylvania," 1960
13.  U.S. EPA Superfund Public Health Evaluation Manual.  Office of
    Emergency and Remedial  Response,  Washington,  D.C., EPA
    540/1-86/060.  1986.
14.  U.S.  EPA, "Development of Advisory Levels for Polychlorinated
    Biphenyls Cleanup,"  OHEA-E-187., 1986.
15.  Wilson, J.T. and J.F. McNaab, "Biological Transformation of Or-
    ganic Pollutants  in Groundwater."  EOS, 64 1983, 505-507.
16.  Wood, P.R., and R.F.  Land and T.L.  Payan. "Anaerobic Trans-
    formation, Transport, and Removal of Volatile Chlorinated Organ-
    ics in Groundwater," in Groundwater Quality (C.H. Ward, W. Giger,
    and P.L. McCarty, Eds), John Wiley and Sons, New York,'NY, 1985.
128     MODELING OF HAZ MAT TRANSPORT

-------
                                     Effects  of  Diffusion  Upon
                       Multi-Dimensional  Contaminant  Transport

                                         Maj Mark N. Goltz, Ph.D., P.E.
                                         Air Force Institute of Technology
                                           Wright-Patterson AFB,  Ohio
                                              Paul V.  Roberts, Ph.D.
                                         Department of Civil Engineering
                                                 Stanford University
                                                 Stanford,  California
ABSTRACT
  Previous experimental work conducted using laboratory soil
columns has demonstrated that diffusion into regions of immobile
water can have a large effect on one-dimensional solute transport
through porous media. Recently, it was shown that the results of
a large-scale field experiment conducted to  study chlorinated
hydrocarbon subsurface transport could be interpreted using a
multi-dimensional transport model  which  assumed that con-
taminant transport was diffusion-limited. The model described
contaminant transport through regions of mobile water using the
traditional advection/dispersion equation. Coupled with this equa-
tion  was a second partial differential equation which described
solute transport within regions of immobile water using  Pick's
diffusion equation. This diffusion-limited model qualitatively simu-
lated experimental results, whereas the simple advective/dispersive
model, that traditionally has been used, could not.
  In this paper, comparisons are made between model simulations
of the multi-dimensional diffusion-limited solute transport model
and the advective/dispersive model. It is shown, using reasonable
parameter values which could be inferred from the field experi-
ment, that the effect of diffusion can be quite large. The diffu-
sion, limited model predicts longer persistence of contaminants at
sampling wells than is predicted by the advective/dispersive trans-
port model. Thus, it is possible that under certain circumstances,
where the diffusion-limitation may be important (e.g. in soils with
low permeability lenses or in aggregated soils), a model which
neglects the diffusion mechanism  can  be  nonconservatively
inaccurate simulating the behavior of contaminant plumes.

INTRODUCTION
  In recent years, investigators have hypothesized that solute diffu-
sion into regions of immobile water could have a significant effect
on  breakthrough response curves observed  in laboratory soil
column experiments 4-8-9> 10' "• 12- Goltz 5 postulated that this
diffusion mechanism could be used  to interpret the results of a
large-scale field experiment conducted to study subsurface trans-
port of chlorinated  hydrocarbons. Bibby1  applied a diffusion-
limited model in a study of chloride movement in a chalk aquifer.
With few exceptions, however, very little work  has been done
regarding field application of these diffusion-limited models. Field
application of these models would appear to be a fruitful  area of
research, especially in light of the laboratory evidence that the
diffusion  mechanism can  significantly  affect contaminant
transport.
  In this paper, simulations of a multi-dimensional model which
incorporates the diffusion mechanism will be compared with
simulations of the advective/dispersive model which commonly is
used to model subsurface contaminant transport. It is shown that
the diffusion-limited model predicts longer persistence of con-
taminants at sampling wells than is predicted by the advective/
dispersive model. Thus, we see that models such as the advective/
dispersive model, which neglect the diffusion mechanism, may con-
siderably underestimate the length of time a contaminant may be
sampled at a well downgradient of a pollution source.

DIFFUSION-LIMITED MODEL
  Goltz and Roberts6 presented the solution to a model  which
postulated advective/dispersive contaminant transport through a
two-dimensional mobile water region with contaminant diffusion
into rectangular layers of immobile water. Figure 1 schematically
depicts the assumptions upon  which such a model is based.
Advection is in the positive x-direction. Longitudinal dispersion
is in the x-direction, transverse dispersion is in the y-direction.
Diffusion is in the z-direction, within the immobile regions. Con-
ceivably, the mobile regions represent high permeability sand layers
                          Figure 1
     Schematic Depiction of Diffusion—Limited Transport Model
                                                                          MODELING OF HAZ MAT TRANSPORT    129

-------
while the immobile regions could represent low permeability silt
layers. Sudicky, et al.," used a similar model to describe solute
transport in a laboratory sandbox constructed using a layer of sand
sandwiched between two silt layers.

COMPARISON OF THE DIFFUSION-LIMITED AND
ADVECTIVE/DISPERSIVE MODELS
   In this section, simulations of the diffusion-limited model will
be compared with simulations of the more commonly used advec-
tive/dispersive model. As stated in the previous section, solutions
to the multi-dimensional diffusion-limited model may be found
in Goltz and Roberts6. Solutions to the multi-dimensional advec-
tive/dispersive model are well known2-3. To compare simulations
of the two models, a set of parameter values is needed. Parameter
values used for the diffusion-limited model are listed in Table 1.
These parameter values were derived  by Goltz5 from the results
of a large-scale field experiment conducted to  study chlorinated
hydrocarbon subsurface transport at Borden, Ontario7. Parameter
values used in the advective/dispersive model, also shown in Ta-
ble 1, are the same as for the diffusion-limited model. Of course,
the advective/dispersive model implicitly assumes that the regions
of immobile water have no effect on contaminant transport so that
parameter values in Table 1  which describe the immobile regions
are not needed for use of the advective/dispersive model.
                            Table 1
                       Model Parameters
    Avenge pore vater velocity*                     O.Otl m/a

    Longitudinal diapereion coefficient"               0.02 pi'/d

    Tranareree dlaperaion coefficient*                0.001 ซ2/d

    Dlffuaion coefficient xlthln Ueaoblle larora        J.lilo"* ป2

    Half-vldth of lUBODila layera                    0.01 •

    Distribution coefficient for tetrechloroethylene*    0.41 eaiVe

    Total water content*                            0.31

    Mobile water content                            0.21

    Soil bulk deneitjr*                              1.11 g/oa1

    Fraction of  aorption altea in the Mobile region      u.40



    •Parameter*  uaed la advectlve/dlaperilve node]
            SIMULATED  CONCENTRATIONS AT A WELL
                          1-20 (Man Y-J maun
  Figure 2 shows breakthrough response curves for tetrachloro-
ethylene simulated by the two models at a sampling well located
20m downgradient and 3m crossgradient from the point of con-
taminant  injection. As  is  apparent from the  figure,  the peak
concentration of contaminant predicted by the advective/disper-
sive model is much greater than the peak contaminant concentration
predicted by the diffusion, limited model. However, the diffusion-
limited model predicts higher concentration levels of contaminant
persisting for years after the advective/dispersive model simulates
negligible concentration levels.
  This  effect may be seen more clearly by examining Figure 3.
Figure 3 plots the logarithm of the ratio of concentration predicted
by the diffusion-limited model to the concentration simulated by
the advective/dispersive model. At early times, the ratio is slightly
negative, indicating that higher peak concentrations are predicted
using the advective/dispersive model. However, due to the long,
Persistent tail of contaminant predicted by the diffusion-limited
model;  over time, the ratio of concentrations rapidly increases. At
the end of 10 years,  the concentration levels  simulated by the
diffusion-limited model are many orders of magnitude greater than
the levels simulated by the advective/dispersive model. In instances
where mg/1 and even pg/1 of contaminant may be of concern,
advective/dispersive model predictions may give a false sense of
security. Similarly, in aquifer  restoration projects, the advec-
tive/dispersive model may  predict  "cleanup" well before actual
cleanup occurs, if the diffusion limitation plays a significant role.

                            Figure 3
              Ratio of Concentrations Simulated by the
 Diffusion-Limited and Advective/Dispersive Models at a Sampling Well

           RATIO OF SIMULATED CONCENTRATIONS
                              TIME (VIMS)
                           Figure 2
          Comparison of Concentrations Predicted by the
         Advective/Dispersive and Diffusion-Limited Models
                             me (vows)

CONCLUSIONS
  Using model simulations, we have shown that the effect of the
diffusion mechanism upon contaminant transport in a hypothe-
sized Held setting can be significant. Neglect of this  mechanism
can lead to underpredicting the length of time during which a con-
taminant may be found at a sampling well. Thus, at  least under
certain conditions, it may be important for models which are used
in an attempt to predict contaminant movement to incorporate the
diffusion mechanism.

REFERENCES
 1.  Bibby, R.,"A Numerical Model of Contamination by Mine Drainage
    Water of the Chalk Aquifer," Tilmanstone, Kent, Water Research
    Centre Report LR 1005, Medmenham Laboratory, Medmenham,
    Marlow, Bucks., Great Britain,  1979.
 2.  Carslaw, H.S. and J.C. Jaeger,  Conduction of Heat in Solids  Ox-
    ford University Press, Oxford, 1959.
130    MODELING OF HAZ MAT TRANSPORT

-------
3.  Crank, J., The Mathematics of Diffusion. 2nd ed., Oxford Universi-
   ty Press, Oxford, 1975.
4.  Crittenden, J.C., N. J. Hutzler, D.G. Geyer, J.L. Oravitz and G. Fried-
   man, "Transport of Organic Compounds with Saturated Groundwater
   Flow:  Model Development  and Parameter  Sensitivity,"  Water
   Resources Res., 22, 1986, 271-184.
5.  Goltz, M.N., "Three-dimensional analytical modeling of diffusion-
   limited solute transport," Ph.D. thesis, Stanford University, Stanford,
   CA,  1986.
6.  Goltz, M.N. and P.V. Roberts, "Three-dimensional solutions for solute
   transport in an infinite medium with mobile and immobile zones,"
   Water Resources Res. 22, 1986, 1139-1148.
7.  Mackay, D.M., D.L. Freyberg,  P.V. Roberts and J.A.  Cherry, "A
   Natural Gradient Experiment on Solute Transport in a Sand Aquifer:
   I. Approach and Overview of Plume Movement," Water Resources
   Res.  22, 1986, 2017-2029.
 8.  Miller, C.T. and W.J. Weber, Jr., "Sorption of Hydrophobic Organic
    Pollutants in Saturated Soil Systems,"/. Contaminant Hyd. 1, 1986,
    243-261.
 9.  Nkedi-Kizza, P., P.S.C. Rao, R.E. Jessup and J.M. Davidson, "Ion
    Exchange and Diffusive Mass Transfer During Miscible Displacement
    Through an Aggregated Oxisol," Soil Sci. Soc. Am. J., 46, 1982,
    471-476.
10.  Rao, P.S.C., D.E. Rolston, R.E. Jessup and J.M. Davidson, "Solute
    Transport in Aggregated Porous Media: Theoretical and Experimen-
    tal Evaluation," So/7 Sc. Soc. Am. J, 44,  1980, 1139-1146.
11.  Sudicky, E.A.,  R.W. Gillham and E.O. Frind, "Experimental Inves-
    tigation of Solute Transport in Stratified Porous Media. 1. The Non-
    reactive Case," Water Resources Res. 21,  1985, 1035-1041
12.  Van Genuchten, M. Th. and P.J. Wierenga, "Mass Transfer Studies
    in Sorbing Porous Media: 1. Analytical Solutions," Soil Sci. Soc. Am.
    J., 40, 1976, 473-480.
                                                                                     MODELING OF HAZ MAT TRANSPORT     131

-------
                          Regulation  of  Dioxin as  a  Carcinogen:
                          A  National  and  International  Dilemma

                                             Norbert  P. Page,  D.V.M.
                                               Dynamac Corporation
                                                 Rockville, Maryland
ABSTRACT
  The weight-of-evidence clearly supports the designation of dioxin
as an animal carcinogen and probably as a human carcinogen. It
appears, however, that the mechanism is one of promotion rather
than initiation. Based upon this, many foreign governments and
some U.S. states have conducted risk assessments for carcinoge-
nicity by the ADI-safety factor (threshold) method rather than the
standard cancer mathematical extrapolation (no-threshold) method
as used by U.S. Federal Regulatory agencies. The result is that U.S.
regulatory agencies have proposed allowable exposure levels only
about 1/lOOth that of other countries.

INTRODUCTION
  Assessing  the   risks  associated   with   exposure   to
2,3,7,8-tetrachlorodibenzo-p, dioxin (TCDD), hereafter referred
to as  "dioxin",  has  been  hotly debated among scientists and
government officials in the United Sates and foreign countries. It
is clear that dioxin is an extremely potent carcinogen to animals.
There is also evidence that workers exposed to dioxin-contaminated
materials may be at greater  risk for  certain types of cancer. Why
then should there be  such a dilemma on regulating dioxin as a
carcinogen? The focus for the continuing debate is quite simple—
the mechanism by which  dioxin causes cancer appears to be one
of promotion rather than initiation. That is, dioxin may not  directly
alter cells to become cancerous (a process known as initiation) but
instead it may simply stimulate already-transformed  and dormant
cells to grow into frank cancers (a process known as promotion).
The end result of either process is an increased risk for  cancer.
Theoretically, however, the promotion process may not occur at
low exposures; that is, there could be a threshold  dose  for the
carcinogenic response. Therein lies  the controversy" With initia-
tion, it is generally accepted that there is no threshold for the effect
and that  any  exposure, no matter  how minute,  could result in
cancer. Government assessments of the carcinogenic  risk of dioxin
have varied greatly depending primarily on whether dioxin has been
declared an initiator or a promoter.
  Virtually all valid  risk assessments incorporate  a weight-of-
evidence approach considering available human data, animal test
results, pharmacokinetics, results of cell studies (e.g., mutageni-
city) and the mechanism for the carcinogenicity. It is the purpose
of this paper to present the weight-of-evidence on which  risk
assessors have relied and  to describe the differing risk assessment
approaches and results.

KNOWN EFFECTS OF DIOXIN TO HUMANS
  Many of the toxic effects observed in animal studies of dioxin
have not as yet been encountered in humans.  Much of our
knowledge of the effects of dioxin to humans has been obtained
from occupational exposures of workers to dioxin in the production
of 2,4,5-trichlorophenol and products made from it.
  Chloracne has been the most consistently observed  symptom,
often the only sign of exposure to dioxin.  Manifested primarily
by chloracne cysts and comedos, chloracne has occurred mainly
on the face, neck and ears. Often chloracne is not an immediate
manifestation of exposure, but may not become evident for several
weeks and once the lesions develop, they may persist for several
years. Other dermatological effects also have been observed in
humans,  including increased pigmentation and hair growth.
  Neurological effects have also been reported including peripheral
neuropathies, and impairments of sight, hearing, smell and taste.
Other symptoms have included nausea, headaches, fatigue, weight
loss, muscular aches and pains, psychological disorders, insom-
nia and a tender and enlarged liver. Clinical studies have revealed
abnormal liver function, altered hematologkal parameters, reduced
peripheral  nerve  conduction,  abnormal  glucose  tolerance,
prolonged bleeding times and the urinary excretion of uropor-
phyrins.
  One  child exposed to dioxin  contaminated  soil in  Missouri
developed hemorrhagic cystitis but  recovered. Domestic animals
died from exposures at the diovin contaminated sites in Missouri
and Seveso, Italy. However, humans,  residing in the same areas
developed minimal effects. It would appear that humans may be
less sensitive to acute toxicity than most laboratory and domestic
animals.  On the other hand, it might be that humans, by their
habits,  simply may not have ingested dioxin in the quantities that
domestic animals did. Stevens' has estimated that the minimum
toxic dose of dioxin for man is about 0.1  Mg/^g-
  So far, there have been no documented deaths, serious illnesses,
immunological changes, birth defects, increases in  spontaneous
abortion and no chromosomal changes attributable to dioxin. The
evidence regarding increased risk for cancer also is conflicting and
will be discussed later.

TOXICOK1NETICS
  There are virtually no data on which to  predict toxicokinetics
of dioxin in humans, and any existing data have been derived from
animal studies. Absorption of dioxin from the gastrointestinal tract
is highly  dependent upon the media it is in. When free in water
or food, approximately 50% may be absorbed. When in ashes or
soil very little is absorbed as the dioxin  is tightly bound  to the par-
ticulate materials. Little is known regarding absorption of dioxin
when inhaled.
132    HEALTH ASSESSMENT

-------
  The metabolism of dioxin appears to proceed via an epoxide
intermediate. The significance of this is that epoxides, in general,
have been associated with DNA damage and binding, which in turn
may initiate carcinogenesis. However, several studies have found
that virtually all the dioxin in liver and adipose tissue of animals
is unmetabolized. When dioxin is metabolized, it appears to be
a rapid process with the metabolites quickly excreted. Thus, little
free epoxide is likely available in the body. In addition, the maxi-
mum covalent binding of dioxin to DNA has been determined to
be four to six orders of magnitude lower than for most chemical
carcinogens.  These data strongly suggest that dioxin metabolites
are rapidly excreted into the urine and bile and that specific cova-
lent binding to tissue  macromolecules does not occur.

EVIDENCE  FOR CANCER IN HUMANS
  Over 20 epidemiology studies have sought to establish a causal
association between exposure to dioxin and cancer; however,  the
results to date have been inconclusive. The major deficiency has
been the great uncertainty as to the existence or magnitude of dioxin
exposure.  Usually the investigators  have only  assumed that
exposure to dioxin has occurred during the use of phenoxy herbi-
cides and chlorophenols or from industrial workplace exposure
(including accidental releases) in trichlorophenol process facilities.
In some cases, especially industrial accidents, the development of
chloracne indicated probable dioxin exposure. Another major
difficulty in determining dioxin exposure is that the level of dioxin
contamination in  herbicides has varied with  the product,  the
manufacturer and over time. In most cases, exposures  have been
reconstructed using  questionnaires which attempted to recall
exposure to  potential dioxin-containing substances. In general,
these potential exposures were to males only and for very short
durations, in some cases only for a few days. In the case of indus-
trial accidents, they were generally an acute, single exposure. A
complicating factor in evaluating the possible effects of dioxin
exposure is that humans have been exposed to other chemicals as
well. Increased incidences of cancer as related to dioxin exposures
have been reported for soft-tissue sarcomas, malignant lymphoma,
stomach cancer, lung cancer and nasal and throat cancer. Most
studies have failed to show an effect; however, the power to detect
an  association has often been inadequate due to the small size of
the study groups.

Soft-Tissue Sarcomas
  The initial observations that cancer might be associated with
exposure to dioxin were made in Sweden, where a 6-fold increase
in soft-tissue sarcomas (STS) appeared to be related to occupa-
tional exposure to phenoxyacetic acid herbicides that  were con-
taminated  with dioxin3'4 In an effort to clarify whether dioxin
could be linked to the increase in STS, the exposures were further
broken down into phenoxy acids without dioxin (i.e., no 2,4,5-T.
etc.) and those that might contain dioxin. The relative risk  for
dioxin-free phenoxy acid exposure was 4.2 versus 17 for exposure
to dioxin-containing phenoxy acids. Thus, it would appear from
this study  that dioxin is  not the only  chemical  that might be
associated with the increase in STS, but that it may indeed play
a substantial role. These Swedish studies have been criticized
primarily on the basis of bias due to selective recall of historical
exposure data by the patients or next-of-kin. They have also been
faulted due to the existence of confounding  factors, such as ex-
posure to chemicals other than phenoxy herbicides or dioxin, and
since the diagnoses of tumors were  not confirmed.
  Many other epidemiology studies involving dioxin exposure have
failed to  identify an increased incidence In the occurrence of STS.
An extensive epidemiologic investigation of Air Force personnel
involved in spraying herbicides,  primarily Agent Orange, in Viet-
nam, did not find an increase in any malignant or benign systemic
tumors4. Two other case control studies of soft-tissue sarcomas
as related to  military  service in Vietnam and potential exposure
to dioxin-containing herbicides failed to find an increase in soft-
tissue sarcomas5'6. It is also apparent that attributing soft-tissue
sarcomas to exposure to dioxin rather than to the phenoxy acid
herbicides  may be  inappropriate.  Considerable evidence  is
mounting that phenoxy acids themselves (or an impurity other than
dioxin) may be associated with various types of cancer, including
soft-tissue sarcomas. In a followup study of cancer among workers
employed in the manufacture of non-dioxin containing phenoxy
herbicides in Denmark, Lynge7 found five cases of STS  in con-
trast to the 1.84 expected.
  It is apparent that the evidence linking dioxin exposure with the
development of soft tissue sarcomas is weak;  nevertheless, the
Swedish results cannot be fully discounted.
Malignant Lymphoma
  The possible association of malignant lymphomas with exposure
to phenoxy  acid herbicides and chlorophenols has  also been
reported by Swedish scientists, claiming a risk ratio of 6.08. This
study also has criticized on the basis of recall bias and confounding
factors including exposure to organic solvents, other chemicals and
smoking. The U.S. National Cancer Institute recently reported a
6-fold increase of non-Hodgkin's lymphoma for men exposed to
herbicides more than 20  days/year and an 8-fold increase for
frequent users who mixed or applied herbicides free of dioxin. The
results suggest that chemicals other than dioxin may be respon-
sible for the induction of malignant lymphomas.

Stomach Cancer
  Two studies reported increased risk for stomach cancers,  one
of workers exposed to dioxin in an industrial accident in 19539
and the other in a cohort of Swedish railroad workers exposed for
at least 45 days to phenoxy acids during 1957-1972 10. The inci-
dence of stomach cancer  cases was  increased 5-6 times that
expected. The number of cases was very small, so  confidence in
the data is limited. No other epidemiologic studies have indicated
an increased risk for stomach cancer. This may be due to the small
size of the cohorts and insufficient power to detect the effect.

EVIDENCE FROM ANIMAL STUDIES
  Four adequately designed and conducted carcinogenic bioassays
of dioxin have been reported. All four indicated clear evidence for
carcinogenicity.  In  1979,   Toth11   reported  that  male
Swiss/H/RiOP mice which were dosed once per week for one year
and observed for their lifetime had  an increase in liver tumors at
0.7 fig/kg/week. As the liver tumors were not identified as benign
or malignant, the data are not amenable to risk assessment.
  The most definitive of the carcinogenicity bioassays was reported
by Kociba, et al., in  19782. Sprague-Dawley rats were  fed a diet
containing dioxin for 2 years at doses of 0.001,  0.01, and 0.1
/tg/kg/day. Statistically significant (p < 0.05) increases in several
types of tumors were reported, including squamous cell carcinomas
of the tongue,  nasal turbinates  and  hard palate in males,  and
squamous cell carcinomas of the nasal turbinates, hard palate, and
lung, and liver  tumors  in  females (Table 1).
  The NTP has conducted two long-term carcinogenicity bioassays
of dioxin, one by dermal exposure3  and the other via gavage4. In
the gavage study, dioxin was administered in corn oil by stomach
tube to Osborne-Mendel rats and B6C3F-1 mice. In  the rat gavage
study dioxin was administered twice per week at doses of 0.0, 0.05
or 0.5 /ig/kg/week for 2 years. As indicated in Table 2, increases
in subcutaneous fibromas and thyroid follicular cell adenomas  in
males and subcutaneous tumors, adrenal cortical adenomas and
liver tumors in  females were  observed. With  the exception of
thyroid follicular cell adenomas, the increases were in the high-
dose groups only.
  In  the  gavage  bioassay with  B6C3F-1  mice, dioxin  was
administered twice per week for 2 years. Increases in liver tumors
in male mice and for subcutaneous fibrosarcomas, histiocytic lym-
phomas, liver tumors, and thyroid follicular -cell adenomas in fe-
male mice were reported (Table 3). All increases were at the highest
dose levels,  0.5 /ig/kg/week for males and 2.0 /ig/kg/week for
                                                                                              HEALTH ASSESSMENT     133

-------
                              Table 1
          Increased Tumor Incidences in Sprague-Dauley Rats
            Administered Dioxin in the Diet for Two Years
Tuner Type/Tissue
                               Dose Level o( TCDD r Type/Tlliue
                                                                       Hapatocellular care in

                                                                       Hapttocellular Carclrt

                                                                         MenoM
                                                                                                                DOM LซYซll
                                                                                                               UJ*-Ooปe  Hid-DOM  X10I-DOM
                                                                                                         4/73

                                                                                                         IVTI
                                                                                n/50-

                                                                                21/M-
      Squaitous Cell Carcinoma of          0/86    0/50     1/50     4/49*

        Nasal Turblnates/Hard Palate

      Squa-ious Cell Carcinoma of Lung      0/86    0/50     0/50     1/49*

      Hepatocellular carcinomas           1/86    0/50     2/50     11/49'

      Hepatocellular Nodules              8/86    3/50    18/50*    23/49'

      Hepatocellular Carcinomas/Nodules*   9/86    3/50    18/50*    34/49'
                                                                       lubcutaneoua Plbrourcom       1/1<     l/M     1/W     5/17*

                                                                       Hl.llocrtlc Lr**>oปe           ป/74     ซ/*>     1/41    14/47*

                                                                       Hepatoolluler Orclnoe*        i/U     2/M     2/41     */47*

                                                                       Hepatocellular Carcinoma/        3/7}     6/50     t/41    11/47-

                                                                        Adenomas

                                                                       Thyroid Polllcular all Menau   0/4ป     1/50     i/ซ7     5/46*
 Adapted fro* Koclba et al.,  (12).

•statistically significant (p <0.05).

4As reported In U.S. EPA (1985).
                                                                  Mซptปd  rroซ nrr. 1M2 (14).

                                                                  *Deปl l*v*
J/50

6/JO'
  Pemlea
   Mapt.d [roo NTP,  (13).

  •stetutlolly ilgnl'lcint (p <0.0i).


 134    HEALTH ASSESSMENT
 7/M-

10/50-
Subcutaneous Ptbrotarcoma
Adrenal Cortical kdenceui
Hepatocellular uoduUa
Hapatoceltular CarclnouaV
uodulei
0/7S
11/73
5/75
5/75

2/50
ซ/>9
1/49
1/49

3/50
4/49
3/50
J/50

ซ/ซ9-
14/46*
12/49*
14/49*

SUPPORTING EVIDENCE FOR CARCINOGENICITY
  Several recent mutation tests with S. typhimurium have failed
to detect a mutagenic response. Early tests in 1972-1973 showed
weakly positive results with S. typhimurium and E. coli but had
serious deficiencies, which rightly diminished the confidence in the
test results. Overall, there is insufficient evidence  for mutageni-
city in procaryotes. With eucaryotic cells, positive results were
reported with Saccharomyces cerevisiae D7. However, the increased
frequencies appear to be related to selective survival of spontaneous
mutants at cytotoxic doses rather  than to mutation induction
  Green, et al., reported a weakly positive increase in chromosomal
aberrations in  rats exposed to dioxin for 3 weeks; however, other
tests performed by Green and several other investigators did not
reveal chromosomal changes. Overall, there is little evidence that
exposure to dioxin induces a clastogenic response in either somatic
or germinal  cells.
  While there  is little evidence that dioxin directly  alters DNA or
induces gene  mutations  or  chromosomal  aberrations, there is
mounting  evidence that  dioxin evokes considerable  epigenetic
activity. This includes damage to various intracellular and nuclei
functions including inhibition of mitosis, interference with chromo-
somal assembly and altered membrane functions.

PROMOTION ACTIVITY
  The studies of Pilot, et al.,5 and  Poland, et al.,6 clearly indi-
cate that dioxin has promotional activity. In Pilot's study, dioxin
was a polenl promoter  for liver tumors in female  rats,  following
a single dose of the initiator, diethylnitrosamine (DEN). On a molar
dose  basis, dioxin was approximately million times more potent

-------
than the known promoter, phenobarbital. Poland investigated the
effect of administering dioxin by dermal application either alone
or following a since skin application of the initiators dimethyl-
benz(a-anthracene   (DMBA)  or  N-methyl-N'-nitro-N-
nitrosoguanidine (MNNG). Dioxin induced a strong promoting
effect with both DMBA and MNNG. Dioxin was about 100 times
more potent in promoting skin tumors than the classical promoter,
TPA, on a molar basis. In these studies, dioxin did not induce
tumors when applied alone.
  Promotion has also been  demonstrated in vitro  studies with
C3H/1OT1/2 mouse embryo fibroblasts. Dioxin was about 10,000
tunes more potent than TPA for promotion17. The evidence is
quite substantial from both in vivo and in vitro studies that dioxin
has a very potent promoting activity, although the mechanism for
the promotional activity is unclear.

SUMMARY OF WEIGHT-OF-EVIDENCE FOR
THE CARCINOGENICITY OF DIOXIN
  Two  Swedish case-control  studies  of  soft-tissue sarcomas
reported a highly significant association of STS with exposure to
phenoxy acids or chlorophenols. Neither study  can clearly point
to dioxin as the causative agent. In fact, one study  indicated an
increased risk for STS as related to exposure to a non-dioxin con-
taining herbicide. Little support can be engendered from other
studies. Even though such support is limited, and with due respect
to the methodological faults of the Swedish studies, the results can-
not be discounted and provide weak but suggestive  evidence for
the carcinogenicity of dioxin.  There is less evidence which in-
criminates dioxm as the-cause of other malignant tumors, although
two studies found an increase in stomach cancers. Until such time
as more definitive studies are reported, the evidence is judged as
insufficient that dioxin is carcinogenic in humans.
   In contrast to the equivocal  evidence in humans, exposure  of
animals to dioxin has clearly resulted in carcinogenic responses in
all adequately conducted tests in rats and mice. Liver tumors were
induced in three studies in which exposure was via ingestion.  In
one study  liver tumors  were  observed in mice dosed  at 0.7
/ng/kg/week. In another study,  a significant increase was observed
in liver tumors in female rats fed dioxin at 0.01 and 0.1 mg/kg/day.
There was also an increase in oronasal and lung tumors in female
rats and in oronasal and tongue tumors in male rats. A third study
found a significant increase in  liver tumors  in female rats dosed
by gavage at 0.5 /t/kg/week. A slight increase in thyroid tumors
was observed in male rats at the 0.05 and 0.5 /t/kg/week. Increases
in liver and lung tumors were found in mice dosed by gavage at
0.5 /t/kg/week for males, and 2.0 /t/kg/week for females. In a der-
mal study, an increase in the incidence of fibrosarcomas in male
and female Swiss mice in which dioxin was applied topically to
the skin three times/week at a dose of 0.001  (males) or 0.005
(females) /t/application. Thus positive carcinogenic responses have
been seen with two species, both sexes, and for  tumors in several
different organs.
   While there is considerable evidence that dioxin possesses sub-
stantial epigenetic activity, the evidence is weak  for a direct inter-
action with DNA and the induction of mutations. Several studies
have conclusively  demonstrated  that  dioxin is a  very  potent
promoter, although the mechanism is unclear.

SETTING PERMISSIBLE EXPOSURE STANDARDS.
   Regulatory officials generally determine allowable  exposures by
two basic approaches depending  upon whether the effect(s)  of
concern are threshold or non-threshold effects. As  illustrated in
Figure 1, with a threshold effect, there is a dose level below which
the effect does not occur. In contrast, with a non-threshold sub-
stance, the effect may occur at even the lowest exposure.
  With threshold effects the method used to determine an allowa-
ble exposure generally has been  referred to as the Allowable Daily
Intake or ADI approach. The ADI is basically the allowable
exposure for continuous long-term exposure. Recently, the U.S.
                     Increasing Dose
                           Figure 1
        Comparison of the Shape of the Dose-Response Curve
           For Threshold and No-Threshold Toxic Effects

EPA has modified the ADI approach and has termed the allowa-
ble exposure as  the  "Reference Dose" or RfD.  For sake of
simplicity, I will refer to the method  as the ADI approach.
  In the ADI approach, an ADI is calculated from the results of
available human or animal  studies by determining the lowest
observed adverse effect level (LOAEL) or the highest exposure level
where there is a no observed adverse  effect level (NOAEL) and
applying safety factors to provide a degree of safety. Based upon
the uncertainty and strength of the human or animal data, the
following safety factors are generally accepted:

• 10-for human variability
• 10-for species variability
• 10-for exposure duration less than chronic
• 10-when LOAEL is used rather than NOAEL

  For example, basing an ADI based upon an excellent human
study would require a safety factor of only 10, whereas basing the
ADI on the NOAEL of a subchronic test would require a safety
factor of 1000 (10x10x10). In addition, regulatory agencies may
include  other safety factors to accommodate for deficiencies in
data.
  In the case of non-threshold effects, several mathematical models
have been employed to extrapolate from human or animal data
to predict the effects that might be observed in humans at very
low exposures. Several models are available; however, the  linea-
rized multistage model has gained the greatest acceptance at this
time. As will be noted later, this model was used by those regula-
tors extrapolating the cancer risk of dioxin.

REVIEW OF GOVERNMENT  RISK ASSESSMENTS
OF DIOXIN
  Many government agencies have conducted risk assessments.
However, due to controversy over the mechanism for carcinoge-
nicity (i.e., promotion) the assessments have varied as to toxic effect
of concern (carcinogenicity or reproductive effects) and methods
used to  assess human risk. The two basic approaches have been
the threshold-ADI Safety Factor approach using carcinogenicity
and/or reproductive effects data or using a nonthreshold extrapo-
lation model using carcinogenicity data only. Table 4 presents  a
summary of the risk assessments which are  described  in the
following paragraphs.
  The U.S. EPA18 concluded that dioxin is a  probable human
carcinogen. Using the multistage linearized model, an incremental
unit risk potency estimate of  1.56xlQ-5  (mg/kg/day)~l was
derived  for dioxin. This was based upon the induction of tumors
in female rats.12 The relative potency index for dioxin is  5 X07
                                                                                             HEALTH ASSESSMENT    135

-------
                           Table 4
            Comparison of Risk Assessments for Dloxln
Country/agency
U.S. EPA


U.S. CDC
U.S. FDA

California
New York


Canada


Netherlands


Switzerland


West Germany


Critical
Effect(s)
Liver cancer


Liver Cancer
Liver Cancer

Liver Cancer
Liver Cancer I
Reproductive
Effects
Liver Cancer I
Reproductive
Effects
Liver Cancer t
Reproductive
Effects
Liver cancer I
Reproductive
Effects
Liver Cancer t
Reproduct ive
Effects
Risk
Assessment
Method
Linearized
Multistage
Model (LMM)
UK
Safety Factor

-------
some U.S. states have used the ADI-safety factor method for the
determination of permissible exposures to  dioxin. This results in
about a 100-fold  increase in the allowable exposure level than if
the standard mathematical models are utilized. Cogent scientific
arguments can be made for both approaches. At this time there
is no clear scientific basis on which to choose one approach over
the other with respect to dioxin. Thus, there does not appear to
be a near term solution to this international dilemma.

REFERENCES
 1.  Stevens, K.M. "Agent orange toxicity: A quantitative perspective."
    Human Toxicol.  1,  1981, 31-39.
 2.  Hardell, L., and Sandstrom, A., "Case-control study: Soft-tissue
    sarcomas and exposure to phenoxyacetic acids or chlorophenols." Br.
    J.  Cancer 39,  1981, 7-77.
 3.  Eriksson, M., Hardell, L., Berg N.O., Moller, T., and Axelson, 0.
    "Soft-tissue sarcomas and exposure to chemical substances: A case-
    referent study." Br. J. Ind. Med. 38, 1981, 27-33.
 4.  Lathrop, G.O., Wolfe, W.D., Albanese, R.A. and Moynahan, P.M.
    "An Lepidemiologic investigation of health effects in air force per-
    sonnel following exposure to herbicides." In: Biological Mechanisms
    of Dioxin Action, (A. Poland and R. Kimbrough, Eds.).  Banbury
    Report 18,  1984, 471-474.
 5. Greenwald, P., Kovasznay,  B.,  Collins, D.N., and Therriault, G.
    "Sarcomas of soft tissues after Vietnam service." J. Natl. CancerInst.
    73, 1984, 1107-1109.
 6. Kang, H., Enzinger, F., Breslin, P., Feil, M., Lee, Y. and  Shepard,
    B. "Soft Tissue Sarcoma and Military Service in Vietnam: A Case Con-
    trol Study." Dioxin 86, Symposium, September 1986, Fukuoka, Japan.
 7. Lynge, E. "A follow up study of cancer incidence among workers in
    manufacture of phenoxy herbicides in Denmark." Br. J. Cancer 52,
    1983, 259-270.
 8. Hardell, L., Eriksson, M., Lenner, P. and Lundgren, E. "Malignant
    lymphoma  and exposure to chemicals, especially organic  solvents,
    chlorophenols, and phenoxy acids: A case-control study." Br.  J. Cancer
    43, 1981, 169-176.
 9. Thiess,  A.M., Frentzel-Beyme, R. and Link, R. "Mortality  study
    of persons exposed to  dioxin in a trichlorophenol-process accident that
    occurred in the BASF AG on Nov. 17,  1953." Am. J. Ind. Med. 3,
    1982, 179-189.
 10. Axelson, 0., Sundell, L., Andersson, K., Edling, C., Hogstedt, C. and
    Kling,  H. "Herbicide exposure and tumor mortality:  An updated
    epidemiologic investigation of Swedish railroad workers."  Scand. J.
     Work Environ. Health 6, 1980,  73-79.
 11. Toth, K., Somfai-Relle, S., Sugar. J. and  Bence, J., "Carcinogeni-
    city testing of herbicide 2,4,5-trichlorophenoxyethanol containing
    dioxin and of pure dioxin in Swiss mice." Nature (London). 278 (5704),
    1979, 548-549.
 12. Kociba, R.J., Keyes, D.G., Beyer, J.E., Carreon, R.M. and Gehring,
    P.J. Long-term toxicologic studies of 2,3,7,8-tetrachlorodibenzo, p-
    dioxin (TCDD) in laboratory animals." Ann. N.Y. Acad.  Sci. 320,
    1979, 397-40.
 13. NTP (National Toxicology Program) "Carcinogenesis Bioassay of
    2,3,7,8-Tetrachlorodibenzo-p-dioxin (CAS No.  746-0-6)   in Swiss-
    Webster Mice (Dermal Study)."  NTP Technical  Report Services
    Issue 20, 3 pp., NTP 80-32;  NIH No.  82-1757, 1982.
14.  NTP (National Toxicology Program) "Carcinogenesis Bioassay of
    2,3,7,8-Tetrachlorodibenzo-p-dioxin (CAS No. 746-0-06) in Osborne
    Mendel Rats and B6C3F1 Mice (Gavage Study)." NTP Technical Report
    Series Issue 209, 95 pp., NTP 80-3; NIH 82-765, 1982.
15.  Pilot, H.C., Goldsworthy, T.,  Campbell, H.A. and Poland, A. "Quan-
    titative evaluation of the promotion by 2,3,7,8-tetrachloro-dibenzo-
    p-dioxin of hepatocarcinogenesis from diethylnitrosamine." Cancer
    Res. 40,  1980, 3616-3620.
16.  Poland, A., Palen, D. and Glover, E. "Tumor Promotion by dioxin
    in skin of HRS/J mice. Nature (London) 300 (5889), 1982, 271-273.
17.  Abernethy,  D.J. Greenlee, W.F., Huband, J.C. and Boreiko, C.J.,
    "2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) promotes the trans-
    formation of C3H/1OT1/2 cells." Carcinogenesis 6, 1985, 651-653.
18.  U.S.EPA'Health Assessment  Document for Polychlorinated dibenzo-
    p-dioxins. Final Report. EPA Report No. EPA-600/8-84-014F, U.S.
    EPA , Office of Environmental Criteria and Assessment, Cincinnati,
    OH,  1985.
19.  Kimbrough, R.D., Falk, H., Stehr, P. and Fries, G. "Health implica-
    tions of 2,3,7,8-tetrachlorodibenzodioxin (TCDD) contamination of
    residential soil." /.  Toxicol.  Environ. Health 14, 1984, 47-93.
20.  Cordle, F. "Use of epidemiology in the regulation of dioxins in the
    food supply." Toxicol. Pharmacol. 1,  1981, 379-387.
21.  California,  State of "Staff report:  Initial Statement of Reason for
    Proposal Rulemaking. Public Hearing to Consider  the Adoption of
    a Regulatory Amendment Identifying Chlorinated Dioxins and Diben-
    zofurans as Toxic Air Contaminants." Air Resources Board, Sacramen-
    to, CA., June 6, 1986.
22.  Kim N. and Hawley, J. "Revised risk assessment, Binghamton State
    Office Building, Albany N.Y." New York State Department of Health,
    Bureau of Toxic Substance Assessment, Jan. 17, 1984.
23.  Murray, F.J., Smith, F.A., Nitschke, K.D., Humiston, C.G., Kociba,
    R. J. and Schewtz, B.A. (1979). "Three-generation reproduction study
    of rats given 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) in the diet."
    Toxicol.  Appl. Pharmacol SO, 1979, 241-252.
24.  Ontario, Province of,  "Polychlorinated dibenzo-p-dioxins (PCDDs)
    and polychlorinate dibenzofurans (PCDFs). Scientific Criteria Docu-
    ment for Standard Development No. 4-84." Ontario Ministry of the
    Environment, Toronto, Ontario, Canada.
25.  Van der Heijden, C.A., Knapp, A.G.A.C., Kramers, P.G.N. and van
    Logten, M.J. "Evaluation of the carcinogenicity and mutagenicity of
    2,3,7,8-TCDD; classification  and standard setting. Report Document
    LCM 300/292. Rijks Institute Voor de Volksqezondheid, Bilthoven,
    The Netherlands, 1982.
26.  Germany, Federal Republic of "Report on Dioxins." Erich Schmidt
    Verlaq, Federal Environmental Agency, Berlin (ISBN 3503 02469 7),
    1984.
27.  Switzerland  "Environmental Pollution due to dioxins and furans from
    commercial rubbish incineration plants." Schriftenr.  Umweltschutz
    No.5, TR-82-0266,1982, pp.  19.
28.  IARC (International Agency  for Research on Cancer), IARC Mono-
    graphs on the Evaluation of  the Carcinogenic Risk of Chemicals to
    Humans. Miscellaneous Pesticides" "IARC, Lyon, France, 30, 1983,
    37-56.
                                                                                                      HEALTH ASSESSMENT    137

-------
             Exposure and  Public  Health  Risk  Assessment  for the
                              Baird  & McGuire Superfund  Site
                                           David E. Burmaster, Ph.D.
                                             Burmaster  & Associates
                                           Cambridge,  Massachusetts
                                              Scott K. Wolff, S.M.
                                       Environmental Risk  Sciences,  Inc.
                                             Sacramento, California
                                              John J. Gushue, Esq.
                                       GHR Engineering Associates,  Inc.
                                          New Bedford, Massachusetts
                                             Brian  L. Murphy, Ph.D.
                                              Gradient Corporation
                                           Cambridge,  Massachusetts
                                           Charles A.  Menzie,  Ph.D.
                                               Menzie & Associates
                                             Westford, Massachusetts
 ABSTRACT
   Under contract at this Superfund site ranked 14th on the National
 Priority List, a complete exposure and health risk assessment was
 prepared as a part of the Rl/FS investigation to assist the U.S.
 EPA in evaluating remedial options at the site. The study area of
 approximately 60 acres posed special challenges: (1) because more
 than  102 HSL  compounds—including 2,3,7.8-TCDD—were
 measured at least once in groundwater, soil, surface water and/or
 sediments/  (2) because certain heavy metals and pesticides
 designated as carcinogens were measured in percent concentrations
 in noncontiguous "hotspot" locations, (3) because 44 percent of
 the site is classified as a wetland and 66 % is located in the 100-year
 floodplain and (4) because the Commonwealth seeks to reopen the
 three public supply wells located within 1,500 ft.
  For the risk assessment, we estimated carcinogenic and non-
 carcinogenic human health effects  from chronic exposures to 53
 "critical contaminants" selected for greatest importance according
 to U.S. EPA methodologies. For  the exposure assessment, we
 modeled six environmental pathways for both a  "No Action,
 Current Conditions" scenario and a "No Action,  Worst Case"
scenario: (1) adults drinking groundwater from the site, (2) children
playing in dry soil at the site, (3) children  playing in muck or
sediments at the site, (4) adults eating fish caught from the Cochato
 River adjacent to the site, (5) adults drinking surface water (after
dilution) downstream from the site and (6) children swimming in
the Cochato River adjacent to the site.
  An essential part of the risk assessment focused on the hetero-
geneous characteristics of the site in terms of contaminants and
public access. This analysis dissected the study area, including the
site and its surroundings, into 11 zones by location and depth before
estimating exposures for each medium, for each pathway and for
 each scenario. With this innovative dissection of the site, citizens
and government officials could pinpoint site-specific concerns and
 understand the need for differential and cost-effective remedies.
 With this  "base case" analysis and with estimates of the effective-
 ness of different remedial technologies, the U.S. EPA has selected
 a combination of soil incineration, pump-and-treat, flood control
 levies and containment technologies in the Record of Decision. The
 remedial plan selected is expected  to cost between $40 and $50
 million.
INTRODUCTION
  Baird & McGuire, Inc., began operating in 1912 as a chemical
manufacturing company along the western bank of the Cochato
River in Holbrook, Norfolk County, Massachusetts 1-3. Sometime
before 1970, the facility was converted to a batching operation,
where herbicides, pesticides, disinfectants, soaps, floor waxes and
solvents were mixed, packaged, stored and distributed. Baird &
McGuire, Inc.,  ceased operations at the site in April 1983.
  Some of the raw materials used in the batching operation (i.e.,
pine oil, fuel oil, coal tar, xylcne, kerosene, heavy aromatic naphtha
and solvents) were stored in tank farm vessels. The chemicals could
be piped from the tanks to the buildings.  Other raw materials,
including pesticides such as  malathion and methoxychlor, were
stored in 55-gal  drums throughout the site.  Some products, (e.g.,
disinfectants, soaps and some pesticides), were blended in "vats"
indoors, while the remaining pesticides arrived at the site pre-
packaged for resale. Finished products awaiting shipment were kept
in the storage building. Figure I shows a plan view of the buildings
and the tank  farm.
                                     Storage
                                 /--^Building
                                            Mixing
                                            Building
                             i Q Farm
                         Figure 1
               Map of Buildings and Tank Farm
138    HEALTH ASSESSMENT

-------
  The earliest waste disposal practices took advantage of the nearby
brooks and wetlands.  Wastes were commonly disposed of in
cesspools or pits east of the buildings and allowed to percolate into
the ground. The disposal systems would often overflow and wastes
would migrate into the brook and wetlands north of the site or
into the low area southeast of the site.

Description of the Site
  The Baird & McGuire property is less than 8 acres. The site,
however, is approximately 20 acres in size and the study area dis-
cussed in this paper includes over 60 acres.
  Baird & McGuire's process buildings, tank farm and office
buildings are situated on a hillside which slopes steeply to the north
and east. The facilities are surrounded by deciduous woodlands.
Relatively flat, wet and poorly drained areas in the central and
eastern portions of the site are characterized by dense, deciduous,
wooded wetland vegetation, except for a 200-ft wide clearing which
has been extensively modified by excavations and fillings.
  Organic soils—in the lowlands and swampy areas—consist of
topsoil,  peat and organic  silt. Most of the site is underlain  by
glacially  deposited, stratified, materials composed primarily of
sands and silt with minor amounts of gravel  and cobbles. These
materials are underlain by a sandy, unstratified glacial till, which,
in turn, is underlain by fractured bedrock at a depth of 20 to
40 ft  below surface.
  Generally, the site is a recharge area for groundwater that leaves
the site as surface water. The porosities and hydraulic conductivities
at the site imply transmissivity values at selected locations in the
range from 3,000 to 72,000 gal/day/ft. While  the groundwater on
the site moves generally towards the Cochato River, the variations
in the thickness of the surficial geological strata and the operation
of remedial interceptor pumps complicate the overall flow pattern.
Two small brooks drain through the study area and flow to the
Cochato  River. Two and one-half miles downstream, the Cochato
flows by the former diversion to the Richardi Reservoir.
  Town Wells #1, #2 and #3 are located  1,500,1,280 and 1,000 ft,
respectively, south-southeast of the Baird & McGuire facility. Well
#3 was closed in the early 1970s due to phenol contamination and
Well #2 was closed at the same time because  of the concern that
continued pumping would induce phenol contamination. In July
1980, the Town closed Well #1 following tests that showed high
concentrations  of  benzene,  chloroform,  1,1-dichloroethane,
1,2-trans-dichloroethylene, 1,1,1-trichloroethylene and tetrach-
loroethylene. When open,  the three wells pumped 1.8 mg/d and
supplied  80 % of the in-town water supply for the Town  of
Holbrook4.
  Land use in the general vicinity of the site is intensely commer-
cial and industrial, with a  residential neighborhood to the south
along South Street. Another residential community is located along
the north shore of Lake Holbrook, approximately 2,000 ft to the
southeast of the site. According to the 1980 Census, the Town of
Holbrook has a "record"  population of 11,140. According to a
1985 Massachusetts Census, 117 people (in the Towns of Holbrook
and Randolph) live within  1,000 ft of the site, another 826 people
live between 1,000 and 2,000 ft of the site and another 9,067 people
live between  2,000 ft  and one mile of the site. Approximately
458 people work regularly within 2,000 ft of the site as well.

Contaminants Measured at the Site
  The U.S. EPA and its contractors  measured chemical  con-
taminants in over 500 samples of groundwater, soil, sediments and
surface water taken from the site. The laboratory measurements
made on these samples revealed the presence of 102 different
compounds found at least once in one or more of the environmental
media.
  In groundwater, lead, naphthalene, total xylenes, arsenic and
2-methylnaphthalene each had highest concentrations above 2,500
/tg/1. In addition, benzidine, ethylbenzene, toluene, total xylenes,
zinc,  acenaphthalene,  anthracene, dibenzofuran,  fluorene,
2,4-dimethylphenol, 2-methylphenol, 4-methylphenol and benzyl
butyl phthalate each had maximum concentrations between 1,000
and 2,500 /tg/1 in groundwater.
  In soils, arsenic, chlordane, 2-methylnaphthalene, fluoranthene
and pyrene each had highest concentrations above 6,000,000 /*g/kg.
Chlordane was present in one soil sample at almost 14 % by weight
and chlordane was present in many other samples in  lower
concentrations. Dioxin  (2,3,7,8-TCDD) was measured in several
soil samples with a maximum concentration of 48 /tg/kg.

SELECTION OF ZONES AT  THE SITE
  A common practice  in many environmental risk assessments
conducted for waste sites is to combine the chemical concentration
data into site-weighted average concentration. These values are then
used as inputs in risk assessments and are meant to represent aver-
age chemical concentrations for an entire site. While this method
of approximation may be defended as a means of representing con-
tamination at sites with homogeneous geology, hydrology and con-
tamination, it   may  not  be  appropriate for  sites  having
heterogeneous characteristics.
  Based on the following criteria, the Baird & McGuire study area
was dissected to capture and represent the "patchiness" of the con-
tamination and the variations in the geology and hydrogeology:

• Homogeneous Terrain and  Ecology:  individual zones were
  chosen to represent areas having similar topographical and eco-
  logical characteristics;
• Potential for  Human  Exposure: each zone was chosen  to
  represent a single set of potential exposure possibilities;
• Potential Remedial  Units: although contamination is found
  almost everywhere at the site, each zone was chosen at  the out-
  set to represent potential remedial units;
• Health Risk: each zone was chosen after reviewing the chemical
  concentration  data and associated toxicity values.
                                          I
      \   Zone4N \    \
            v-
                             \
                           I
      /    V*
ne4S  /         ^  ^
    f     Zone 6
    J                Town
    I   T             Well
       To*"    Town     ป1
   I   T,"    We2"      ฐ
                      ^r-> L-    ---^	    x
       /     ' ^VC^C"  V^           Zones  /
       / zone 3 J   ,.  ^    p-	-~—^_y

        \   I     Zonel   /
                           Figure 2
                       Map of 11 Zones
  Based on these criteria, the study area was divided into 1 1 Zones
for separate evaluation and analysis1 ( Fig. 2):

• Zone 1, an upland portion of the Site, contains the abandoned
  tank farm and the abandoned buildings used for offices, storage,
  mixing and laboratory facilities;
• Zone 2, fenced and capped after  the EPA removed approxi-
  mately 1,000 yd3 of contaminated soils in 1984, contains the
  primary former on-site disposal area;
• Zone 3 contains the ephemeral stream called the  Unnamed
  Brook;
• Zone 4N (N for North) contains low-lying areas and wetlands
  along the Cochato River downstream of Zone 2 to Mear Road;
• Zone 4S (S for South) contains low-lying areas and wetlands
  along the Cochato upstream of Zone 2;
• Zone 5 contains an area of suspected overland flow along a dirt
  access road from Zone 1  towards Town Well 3;
• Zone 6 represents the Cochato River and a narrow strip of wet-
  land along each side from the southerly (upstream) extent of the
                                                                                             HEALTH ASSESSMENT     139

-------
   security fence;
 • Zone 7 represents areas across the Cochato from Zones 2, 3 and
   4N (this Zone has no distinct eastern boundary);
 • Zone 8 represents the area to the south of the main site, which
   contains closed Town Wells 1, 2 and 3 (this Zone has no dis-
   tinct southern boundary);
 • Zone  "Up" represents the upstream portion of the Cochato
   River from the dam/weir at Lake Holbrook to Zone 6; and
 • Zone "Down" represents the Cochato River downstream from
   Zone 6 (this Zone also has no distinct downstream boundary).

   These 11 Zones subdivide the study area completely, i.e., no
 portion of the study area remains unzoned and they form the areal
 basis of all the  estimated human health effects.

 RISK ASSESSMENT
   The risk assessment  methodology used in this analysis closely
 followed guidance published by  U.S.  EPA5'7. In addition, the
 Agency's staff members in Region I helped evaluate and choose
 the site-specific exposure factors  for this study. The following
 sections summarize the results:
                          Table 1
           Critical Contaminants Selected for (be Study
          Critical
       Contaminant
      1,1-dichloroethylene
        1,2-dichloroethane
            2,3.7.8-TCDD
               4.4'-DDO
               4,4'-DDE
               4.4--DDT
                  aldrin
                 arsenic
               benzene
               benzidine
          benzo(a)pyrene
               beryllium
              BHC-alpha
               BHC-beta
              BHC-delta
             BHC-gamma
               cadmium
              chlordane
              chloroform
                dieldrin
              heptachlor
       heptachlor epoxide
                  nickel
       tetrachoroethylene
         trichloroethylene
            vinyl chloride
 1,2-trans-dichloroethylene
1,3-trans-dichloropropylene
             2-butanone
                 barium
           ethylbenzene
            lluoranthene
                  lead
                  silver
                toluene
            xylenes, total
                  zinc
        Total Other PAHs
            dibenzoluran
Health
Rating*
  C
  C
  C
  C
  C
  C
  C
  C
  C
  C
  C
  C
  C
  C
  C
  C
  C
  C
  C
  C
  C
  C
  C
  C
  C
  C
 NC
 NC
 NC
 NC
 NC
 NC
 NC
 NC
 NC
 NC
 NC
 SC
 SC
 •Notes: C - Carcinogen; NC - NonCarclnogen; SC - Suspected Carcinogen

140    HEALTH ASSESSMENT
 Hazard Identification
   In sufficiently large doses, each of the 102 compounds measured
 at the site can cause human health effects from either long-term
 (chronic) exposures or short-term (acute or sudden) exposures.
 However, the compounds  have widely differing potencies  in
 mammalian and  human  physiologies. Of the 102 compounds
 measured at the site, dioxin has the highest potency. At the other
 end of the spectrum, zinc, for example, is toxic to humans only
 in extremely large doses. In smaller doses, zinc is an "essential"
 nutrient for humans and many pharmaceutical companies include
 zinc in "vitamin and mineral" tablets.

 Dose-Response Assessment
   Using Cancer Potency Factors (CPFs) and Acceptable Intake,
 Chronic values (AICs) derived and supported by the U.S. EPA's
 Carcinogen Assessment Group and Environmental Criteria and
 Assessment Office, this risk assessment focused on the effects of
 "long, term, low-level" exposures effective for several years to a
 lifetime.
   An examination of the complete data sets for all the measure-
 ments at the site revealed that the 102 contaminants are present
 in widely varying concentrations and spatial distributions. To focus
 the risk assessment on the most important compounds, a list of
 53 "Critical Contaminants" was developed using methods suggest-
 ed by the U.S. EPA.  Basically, the fully quantitative selection
 process ranked compounds according to the product of: (1) their
 maximum concentrations in groundwater and soil samples, (2) their
 toxicity and (3) an exposure factor for each medium. Through this
 procedure, we selected 53 Critical Contaminants. This list later was
 reduced to 38 named compounds and one large category of total
 (other) PAH compounds to group the 15 other polycyclic aromat-
 ic hydrocarbons. As shown in Tables 1 and 2, the final list of Crit-
 ical Contaminants includes 26 compounds considered carcinogens
 by U.S. EPA,  1 noncarcinogen and many suspected carcinogens.

                          Table 2
   PAH Compounds Included in the "Total  Other PAHs" Category

                 2-methylnaphthalene
                 acenaphthene
                 acenaphthylene
                 anthracene
                 benzo(a)anthracene
                 benzo(b)fluoranthene
                 benzo(ghi)perylene
                 benzo(k)fluoranthene
                 chrysene
                 dibenzo(a,h)anthracene
                 fluorene
                 indeno(l ,2,3-cd)pyrene
                 naphthalene
                 phenanthrene
                 pyrene
Exposure Assessment
  For the Base Case conditions at the Baird & McGuire site, poten-
tial exposures of people to different environmental  media were
modeled to consider both:

• A "No Action, Present Conditions"  scenario  thought  to
  represent reasonably well the current conditions at  the site and
• A "No Action, Worst Case" scenario, thought to  represent
  maximum site-specific exposures to hypothetical  future con-
  ditions with no remediation.

-------
  These exposure scenarios form the basis of all subsequent risk
estimates. (See also Table 3 for a summary.)
                            Table 3
                     Exposure Factors Used
            for the "No Action, Worst Case" Scenarios

  Ground Water:

  A 70-kilogram Adult drinks 2 liters of water per day for 70 years.

  Dry Soils:

  A 43-kilogram Child plays on the site for 60 days per year for 14 years. The
  child ingests 100 milligrams of soil per visit. An estimated 0.5 milligrams per
  square centimeter of dry soil covers 1,328 square centimeters of skin (both
  hands and both forearms) during each visit. By assumption, 10 percent of the
  organic compounds and 1 percent of the heavy metals in the soils on the skin
  penetrate the dermal barrier.

  Muck:

  A 43-kilogram Child plays for 30 days per year on the site for 14 years. The
  child ingests 100 milligrams of muck per visit. An estimated  40 milligrams  per
  square centimeter of muck covers 1,374 square centimeters of skin (both hands
  and both feet) during each visit. By assumption, 1 percent of the organic
  compounds and 0.1 percent of the heavy metals in the muck on the skin
  penetrate the dermal barrier.

  Fish:

  A 70-kilogram Adult consumes 250 grams of fish per week for 70 years. For
  organic contaminants, the ratio of concentrations of compounds in the fish to the
  sediment equals 0.77; metals were not calculated.

  Surface Water via Drinking Water:

  Same as for Ground Water (above), except the Cochato provides a factor of 10
  dilution by the point of ingestion.

  Surface Water via Swimming:

  A 43-kilogram Child swims in the Cochato River for 36 days per year on the site
  for 14 years. The child swims for 30 minutes each visit. The child inadvertently
  ingests 50 cubic centimeters of river water is  per visit. Approximately 10,485
  square centimeters of the skin are immersed (90 percent of the body area), and
  organic chemicals penetrate according to a model in Brown, H.D., D.  Bishop,
  and C. Rowan, 1984, The Role of Skin Absorption as a Route of Exposure for
  Volatile Organic Compounds (VOCs) in Drinking Water, Am. J. of Pub. Health.
  74:479-484.
Exposures to Groundwater
   At present, there are no public or private drinking water wells
within 1 mile of the site, but, with no remedial actions taken, people
may drink water present on or migrating from the site. Thus, for
the "No Action, Present Conditions" scenario, we assume that
there is no consumption of this water at present. For the "No
Action, Worst Case" scenario, we assume that a healthy adult
weighing 70 kg ingests 2 I/day for a lifetime.  The risk estimates
presented in the next section are based on the average concen-
trations of contaminants in Zones 1, 2, 3, 4N, 4S, 5, 6, 7 and 8.

Exposures to Dry Soils
   Starting in 1986, most of the site was secured with a chain, link
fence topped with barbed wire. For the "No Action, Present Con-
ditions" scenario, we assume that there is no contact with dry soils
at the site,  but for the "No Action,  Worst Case" scenario, we
assume that contact  may occur.
   Soils at the site have a range of depth to bedrock from a few
ft to over 30 ft. Due to the stratification of the natural soils and
of the contamination, we consider "Surface Soils" (0 to 4 ft deep)
and "Deep Soils" (more than 4 ft deep) separately in all calcula-
tions.  Broadly, we model children, ages  5 through 18, as the
exposed population that may play at the site some days during the
spring, summer  and  fall. These children are exposed to  con-
taminants in this scenario by inadvertent ingestion of soils (by hand
to mouth) and by dermal penetration from the soil matrix.
Exposures to Muck
  These scenarios are based on the concept that children ages 5
through 18 may visit the site to play, e.g., catch frogs. The basic
assumptions on mechanisms strongly parallel those for exposures
to dry soils, although the values of certain exposure factors differ
appropriately.

Exposures to Contaminated Fish
  Fish in the Cochato River come into contact with the sediments
there and bioconcentrate the contaminants to differing degrees.
At present, no fishing is reported in the Cochato River directly
adjacent to the site, however, fishing does occur downstream in
Sylvan  Lake,  which is  hydrologically  connected to  the river.
Broadly,  we assume that healthy adults consume tainted fish
regularly for a lifetime and the frequency of ingestion depends on
the scenario.

Exposures to Surface Water via Drinking Water
  At present, no one is known to sustain a dose of contaminants
from the site  through a surface water  pathway because public
authorities have severed the connection between the Cochato River
and drinking water supplies downstream and because the shallow,
sluggish Cochato River  does not attract swimmers  or boaters.
However, in the future, it is possible that persons could drink water
from the river, possibly diluted, thereby sustaining a dose  of
contaminants.

Exposures to Surface Water via Swimming
  For the "No Action, Present Conditions"  scenario, we assume
no swimming due to the  shallow depths  and  uninviting character
of the water. However, for the "No Action, Worst Case" scenario,
we consider that children, ages 5 through 18,  swim regularly  in
the river. To the extent that children do swim in the Cochato River,
we estimate the dose received by them from (1) inadvertent ingesta-
tion of surface water and (2) absorption of organic contaminants
through the intact skin.

Exposures Not Considered for the Base Case  Conditions
  The Feasibility Study did not make quantitative estimates for
exposures to vapors and fugitive dust transported in air because
no air quality data were collected at the site. The Feasibility Study
also did not estimate human exposures to contaminants infiltrating
into a 12-in. cast-iron water main traversing the site because the
pipe is under pressure and because the Town removed the pipe from
service  in 1986. Finally, the Feasibility Study made only qualita-
tive estimates of potential exposures during remedial activities, i.e.,
excavation of dry soil, wet soils and muck, because this topic would
be re-opened after the selection  of the remedial plan.

Risk Characterization
  In the risk characterization section of the study, we combined
information obtained from the dose-response assessment and the
exposure assessment and made quantitative estimates of the poten-
tial human health effects resulting from chronic exposures to the
53 Critical Contaminants. In a highly disaggregated methodology,
the average concentrations  for each of the contaminants in  each
environmental medium in each of the 11 zones were combined with
the toxicological potencies and the assumed frequencies and magni-
tudes of exposures to make quantitative estimates of human health
risks. The disaggregated results were then combined  to estimate
the potential human health risks for each scenario.
  The  risk characterization process proceeded as follows:

Step 1: The average concentration of each  chemical  in  each
        medium and each zone was calculated from  the labora-
        tory data.
Step 2: The average concentration of each contaminant reaching
        humans  was estimated (1) as the average environmental
        concentration, (e.g.,  for ingestion of groundwater  as
        drinking water),  or  (2)  as the  output of a simple
        model,(e.g., for consumption of fish tissue in chemical
                                                                                                  HEALTH ASSESSMENT    141

-------
        equilibrium with sediments in the stream).
Step 3: The average daily dose  for each  contaminant in each
        zone for each scenario was estimated using the concen-
        trations from the previous step and from the exposure
        factors presented in Table 3.
Step 4: The potential human health effects for each contaminant
        were estimated based on (1) the Cancer Potency Factors
        (CPFs)  and (2) Acceptable Daily  Intake (ADI) values,
        Acceptable Intake-Chronic (AIC) values or the Reference
        Doses (RfD) value published by the U.S. EPA. Contami-
        nants for which the Agency has not provided  toxicologi-
        cal  potencies were not analyzed in this step.
Step 5: Potential chronic human health effects were estimated by
        calculating (1) the incremental lifetime Cancer Risk from
        lifetime exposure to carcinogens and (2) the Health Index
        for  chronic exposure to noncarcinogens. Carcinogenic risk
        is the incremental probability of  a  person manifesting
        cancer sometime during  a lifetime  from  exposure.  As
        defined by the U.S. EPA, the Health Risk is the ratio of
        the actual daily dose to the acceptable daily  dose.
Step  6: In  the  absence  of information  on  possible  synergisms
        and/or  antagonisms, the estimated  Cancer  Risks were
        summed for each pathway  and each scenario. A similar
        procedure was used to combine the  Health Indices.

   The results obtained from the risk assessment represent the com-
pilation of  S3 spreadsheets concerning the 6  exposure routes for
the 11  zones (not every zone considered each exposure route).
Tables 4 and 5  present the estimates for the "No Action Worst
Case" Scenario. Note that all the zones have at least one pathway
with the estimated incremental lifetime cancer risk greater than one
in 1,000 and that all but  two  of the zones have at least one path-
way with the Health Index greater than one.

                           Table 4
   Summary of Estimated Incremental Lifetime Carcinogenic Risks
                No Action, Wont Case Scenarios
 Zone
  1

  2

  3

 4N

 4S

  5

  6

  7

  8

Down
 Ground    Playing in   Playing In    Rsn
 Water     Surface    Mick and   Ingestlon
Ingesllon    Soils    Sediment
byAduts  by Children  by Children   by Adults
                  2 1E-06

                  48E-05
  Drinking    Swimming in
  Diluted      Surface
Surface Water    Water
 by Adults    by Children
50E-04
7.3E*00
4.2E-01
1 4E-03
4.8E-03
3.9E-03

28E-02
6.5E-05
6.8E-04
4.3E-04
8.2E-OS
6.0E-05
46E-08
1.3E-02

49E-03
         3.5E-08
                  1.7E-04
                          1.0E-04
                           5.7E-02
                                   S8E-03
                                    84E-04
                                               5.8E-05
RECORD OF DECISION
   On September 30,1986, the Regional Administrator of U.S. EPA
(Region I) signed the "Record of Decision" (ROD) for the Baird
& McGuire site8.  In  this decision, based  strongly on the risk
assessment,  the U.S. EPA chose a plan to:

•  Excavate in "hot areas" to remove approximately 191,000 yd3
   of contaminated soils
•  Treat contaminated soils utilizing on-site thermal destruction,
   after a test burn and air quality modeling
•  Extract groundwater and treat it at an on-site treatment plant,
   with discharge to the extent feasible to the aquifer on-site

142    HEALTH ASSESSMENT
                                                             Zone



                                                              I

                                                              2

                                                              3

                                                             4N

                                                             4S

                                                              s

                                                              e

                                                              7

                                                              8

                                                            Down
                                                    Tables
                                    Summary of Estimated Chronic Healcb huUcei
                                  Ground   Playing In   Playing in             Drinking    Swimming in
                                  Water     Surface   Muck and     Rah      Oitded      Surface
                                 Ingeillon     Soil*    Sediment   Ingctllon   Surface water     Water
                                 byAduls   by Children  by Children  by Adults    by Adults    by Children
                                  1739

                                  1751

                                   851

                                 28206

                                   4.25

                                  1707



                                  1453

                                   t 88
                    0.06

                    0.29

                    3.1*

                    4.59

                    0.11

                    083

                    002



                    0.05
0.01

005
005
                                                                                        009
         096
                                                              003
                             0 15
  Restore the wetlands where contaminated soils are excavated
  Construct temporary levees for flood protection
  Relocate the unnamed brook
  Monitor the groundwater on- and off-site
  Monitor air quality during remedial construction and implemen-
  tation of the thermal destruction.

  U.S. EPA Region I anticipates that the remedial actions will be
completed circa 1994 (incineration) and circa 1999 (groundwater
treatment) and that the complete remedial program will  cost
between $40  and  $50 million.

CONCLUSION
  The methodology for the health risk assessment and overall
potential human health risks has been presented for the Baird &
McGuire NPL site in Holbrook, Massachusetts. The risk assess-
ment was prepared following guidance from the U.S. EPA. The
large study area includes a complex terrain that—until recently-
presented a wide range of possible human exposures at different
locations. To capture its diversity and heterogeneous characteris-
tics, the study area was divided into 11  zones having different
characteristics in  terms of topography, the  potential for human
exposures and possible remedial alternatives. Potential human
health risks were estimated for each environmental medium in each
zone. The division of the study area into the 11 Zones allowed
decision makers to understand the differential health risks posed
by different portions of the site. In the Record of Decision for the
site, the U.S. EPA recommended remedial alternatives based on
this zonal approach.

REFERENCES
1.  OHR Engineering Associates, Inc., "Final  Feasibility Study Report:
   Baird & McGuire Site, Holbrook, Massachusetts." July 18, 1986.
2.  NUS Corporation, "Preliminary Site Assessment of Baird & McGuire,
   Inc. Holbrook, Massachusetts:" TDD No.  FI-82I2-07, April 1983.
3.  NUS Corporation, "Remedial Action Master Plan.  Baird & McGuire
   Site,  Holbrook, Massachusetts,"  U.S.  EPA Work  Assignment
   No. 01-1V31.0. May 1983.
4.  Special Legislative Commission on Water Supply, Commonwealth of
   Massachusetts, "Water Quality Issues  in Massachusetts:  Chemical
   Contamination." Oct.  1981.
5.  US EPA, "Risk Assessment Guidelines." draft, Fed Reg 49, Nov. 23,
   1984, 46304.
6.  US EPA, "Superfund Exposure Assessment Manual," draft  Janu-
   ary 14,  1986.
7.  US EPA, "Superfund Public Health Evaluation Manual," draft 1985.
8.  US EPA, Region I,"Record of  Decision:" Remedial Alternative
   Selection, signed by Regional  Administrator Michael  R  Deland
   Sept. 30, 1986.

-------
               The Human  Health  Risks  of  Recreational  Exposure
                             To Surface  Waters Near NPL Sites:
                                   A Scoping Level  Assessment

                                                  Gary K. Whitmyre
                                                    James J. Konz
                                                   Mark  L. Mercer
                                                    H. Lee Schultz
                                                      Versar, Inc.
                                                 Springfield, Virginia
                                                    Steve Caldwell
                                      U.S. Environmental Protection Agency
                                          Hazardous Site Control Division
                                                  Washington,  D.C.
ABSTRACT
  An assessment was performed to determine whether exposure
to toxic Superfund pollutants via recreational use of surface waters
near NPL sites poses a significant risk. The first step of this
assessment involved calculation of doses of pollutants received near
NPL sites as a result of swimming and fishing; this calculation was
made using newly  developed  exposure algorithms and actual
monitoring data for 27 pollutants at 12 NPL sites. These doses
were used to calculate risks of cancer and non-cancer effects, which
were summed  across  chemicals  and  activities (i.e.,  fishing,
swimming) and compared  to appropriate target  risk levels to
determine their significance.
  Five (45%) of the 12 example sites have cancer risks that exceed
the 10~6 target risk and three of the sites (27%) exceed the  10 ~5
target risk; these risk numbers do not include the risk from con-
sumption of recreationally caught fish, which would lead to higher
risks. These risks were compared to risks from consumption of
drinking water from surface water sources near the same  sites.
Health-based benchmarks that are expressed as surface water con-
centrations that correspond  to   10 ~6  risk  from recreational
exposures  were developed. This study suggests that recreational
use of surface waters near NPL sites can be a significant source
of risk.

INTRODUCTION
  Numerous mechanisms for release of hazardous pollutants from
Superfund sites into surface water, including drainage discharge,
runoff and interconnection of contaminated groundwater and
nearby surface waters, can result in transport of pollutants to
recreationally used surface waters.1 Recreational exposure to
chemcials released from Superfund sites can result from several
activities including swimming and fishing; these two activities are
expecyted  to have a highparticipation rate.
  Swimming can lead to whole-body dermal exposure to pollu-
tants in surface water, inhalation exposure to pollutants  that
volatize from the water and exposure due to inadvertant ingestion
of surface water. Fishing can lead to partial-body dermal exposure
to pollutants in surface water.
  Dermal exposure considerations were included in the assessment
of risk for swimming and fishing because at  least one study has
shown that dermal contact with water containing volatile organic
compounds (VOCs) can lead to significant dermal exposure.2 Ex-
posure to pollutants ingested during consumption of recreationally
caught fishis thought tocontribute significantlyto risk but was not
addressed  in detail  in this  study. No single study to date has
adequately analyzed the risks of exposure to toxic pollutants from
Superfund sites via recreational use of contamintaed surface waters.
  The study reported here was designed to provide an integreated
assessment of risk due to recreational exposure. A four-step proce-
dure was used to perorm a risk assessment of 27 chemicals at 12
NPL sites. First, an exposure assessment was performed to con-
vert surface water concentrations to time-weighted average (TWA)
daily doses in mg/kg-day. This exposure assessment involved the
development of new exposure algorithms specifically for this study.
Second, risks were calculated for recreational exposure to  each
chemical for each appropriate exposure route (dermal, inhalation
and ingestion). Third, these chemical-specific risks were summed
within  each activity  (fishing, swimming). Fourth, risk  for the
maximum exposed individual (MEI), which is a hypothetical per-
son who both fishes and swims,  was calculated by summing the
risks for the two activities for each site.
  While this study focused primarily on cancer risks which were
calculated using a linear extrapolation model, it addressed  risks
of noncancer chronic effects as well. Recreational risks were also
compared with the risks of consuming the same surface waters as
drinking water; this was done by developing site-specific risk ratios.
Health-based benchmarks for recreational exposure that are surface
water concentrations for each of the 27 chemicals corresponding
to 1  x 10~6 health risk also were calculated.

SELECTION  AND DESCRIPTION OF TEST SITES
  The 12 sites selected for the risk assessment were chosen based
on the  availability of concentration data for adjacent or nearby
surface water(s) and known or suspected recreational use of the
subject surface water. As of Dec.  29, 1986, 572 (64%) of 888 NPL
sites were identified via the HRS data base as sites with potential
recreational use.3 However,  the HRS data base contains only
preliminary information concerning NPL sites. Therefore, for the
purposes of this study, it was necessary to obtain data (such as
the identities of the actual contaminants present and their concen-
trations in surface waters) for sites where the existence of surface
water contamination and recreational  use has been  verified. To
obtain this information, a review was conducted of all RODs avail-
able at  the time of this study. A total of 179 RODs from the U.S.
EPA Headquarters Library, the Superfund Docket Office and
additional staff offices were reviewed. Particular attention was
given to the reporting of surface water concentration data in the
RODs. In some of the RODs, ponds or lagoons located on-site
were classified as surface water; however, because the focus of this
study was off-site surface water use, only sites with data from iden-
tified water bodies off-site that might be used recreationally were
selected.
                                                                                         HEALTH ASSESSMENT    143

-------
  These sites represented a cross-section of types and sizes of water
bodies that usually were immediately adjacent to or within a few
hundred yards of the site. Of these surface waters near the 12 sites,
one represented a small lake, one represented a large lake, two
represented wetlands areas, one represented a major river adja-
cent to the site and the remaining seven represented streams adja-
cent to or near the sites. The distribution of recreational activities
in these surface water bodies was as follows: unspecified water-
related recreation—two;  swimming-, one; fishing—four (two of
these  areas are currently stocked with trout or were at one time);
and suspected recreational use of surface water—five.

SELECTION OF CHEMICALS
   The chemicals chosen as pollutants for this study were selected
based on the frequency of their occurrence at proposed and final
NPL  sites, type of toxicity (emphasis on  cancer or serious organ
effects), availability of unit risk factors, availability of skin perme-
ability constants (for assessment of the dermal route) and availabil-
ity of monitoring data for the pollutants at the sites selected above.
The 27 chemicals chosen for this study are shown in Table I. These
selected chemicals represent a cross-section of different chemical
classes  (including  pesticides, metals,  aromatic  hydrocarbons,
halogenated aromatics, halogenated aliphatics, ketones and esters)
and a wide range of toxic potencies (unit risks varying over six
orders of magnitude).
                           Table 1
            Chemicals Selected for Risk Assessment of
                     Recreational Exposures
   Acetone
   Aldrln
   Barium
   Benzene
   Beryllium
   Chlorobenzene
   Chloroform
   Chloromethane
   Chromium
   OEHP
   DDT
   1,2-Dlchloroethane
   1,1-Dlchloroethylene
   Dleldrln
Ethylbenzene
a-Hexachlorocyclohexane
Llndane
Mercury
Methylene Chloride
Napthalene
Nickel
PCBs (Arochlor  1248)
Phenol
Tetrachloroethylene
Trlchloroethylene
Vinyl Chloride
Xylene
METHODOLOGY
  Each approach used to calculate dose, risk, risk ratios and health-
based benchmarks for recreational surface waters is summarized
below.

Calculation of Dose—Dermal Exposure
  Pollutants released to surface water near NPL sites are expected
to be absorbed into the body both during swimming and fishing.
The equation used to relate the surface water concentration of a
pollutant to the dose received via dermal exposure during swimming
and fishing is as follows:
   =  C
where
  d   =
  C   =
  T   =
  f   =

  P   =
  A   =
        xTxfxJ-xpxAx   ye"
                    bw              365 days
                         I  liter
                       1000cm'
                                                         (1)
        time-weighted average dose (mb/kg-day)
        concentration of pollutant in surface water (mb/1)
        duration of each recreational event (hr/event)
        annual frequency of recreational events (events/year)
        body weight (70  kg)
        dermal permeability constant for pollutant (cm/hr)
        surface area of exposed  skin (cm2).
                                      The last factor in Equation 1  accounts for the estimation that
                                    a person swims only 20 years out of a 70-year lifetime or fishes
                                    40 years out of a 70-year lifetime. Dermal permeability constants
                                    were available for only six of the 27 chemicals assessed in this study;
                                    thus, the dermal component of exposure was assessed for only these
                                    six chemicals. The values for these constants are shown in Table 2.
                                    Table 3 presents the assumed values for various input parameters
                                    (such as duration and frequency of exposure) needed to construct
                                    the exposure scenarios. When exposure parameter values from
                                    Table 3 were inserted into Equation 1, the following relationships
                                    were obtained:
                                    Dermal dose from swimming  =  (C x  p) +  76            (2)

                                    Dermal dose from fishing    =  (C x  p) -r  170            (3)


                                                                Table 2
                                                Dermal Permeability of Chemicals Selected
                                                          For Risk Assessment


                                   Chemical   Dermal Permeability Constant  (cm/hr)   Reference
Aldrln
Dleldrln
Llndane
Benzene
Ethylbenzene
Phenol
5.2 x
5.1 x
6.2 x
4.1 X
1.0 x
8.22
10-ซ
10"ซ
10-ซ
10-1
10-3
X 10-2
(4)
(4)
(4)
(5)
(6)
(7)
                                                                                                Tabte 3
                                                                                       Exposure Parameters Used in
                                                                                       Recreational Risk Assessment
                                      Expoaura    Expoaure  Parameter
                                                                                                            Parameter  Valua
                                                                                                        Swimming       Fishing
                                      Darmal     Duration et Exposure
                                                 Frequency  of Exposure
                                                 Exposed Skin  Area
                                                 Body Weight

                                      Inhalation   Duration of Exposure
                                                 Frequency  ol Exposure
                                                 Inhalation  Rate
                                                 Body Weight

                                      Ingestlon    Duration of Exposure
                                                 Frequency  of Exposure
                                                 Ingestlon Rate
                                                 Body Weight
                                                                                                       2.6 hrs/event*   5.0 hrs/event*
                                                                                                       25 events/yr*   19  events/yr1
                                                                                                       18.150cmปc     2,800cm*>
                                                                                                       70 kg*         70  kg

                                                                                                       2.6 hrs/event   5.0 hrs/event
                                                                                                       25 events/yr    19  events/yr
                                                                                                       2.6 m*/hrh      U mVhr1
                                                                                                       70 kg          70  kg

                                                                                                       2.6 hrs/event      _k
                                                                                                       25 events/yr       I
                                                                                                       50 ml/hrl
                                                                                                       70 kg
                                     •This Is an  average from Reference (8).
                                     b Upper bound limit from Reference (9); due to significant  regional
                                       variation In frequency of outdoor swimming,  a representative
                                       average value could  not be calculated.
                                     c This corresponds to whole body exposure.
                                     d Average adult body weight used for standard rlak assessment
                                     • Average from Reference (10).
                                     ' Average from  Reference (10).
                                     • Assumes exposure of both lower legs and hands.
                                     h Assumes alternating periods of moderate to heavy activity and
                                       reatlng; Inhalation rates from Reference (11).
                                     I  From Reference (11);  corresponds to Inhalation for average adult
                                       performing  light activity.
                                     I  Versar estimate.
                                     * No significant Ingestlon exposure Is expected during fishing.
144    HEALTH ASSESSMENT

-------
Calculation of Dose—Inhalation Exposure
  The methodology for assessment of risk from inhalation ex-
posure to pollutants volatilizing from contaminated surface water
was based on a release algorithm developed to assess exposures
at Superfund sites'; a mixing zone of 2 m in height was assumed.
The mass flux dilution factors derived from this algorithm are
chemical-specific, being partly a function of molecular weight and
Henry's law constant. Because of the inability to obtain Henry's
law constants for some chemicals (e.g., metals), the inhalation com-
ponent of exposure was addressed for only 22 of the 27 chemi-
cals. Values for  mass flux dilution factors for the chemicals
addressed in this  risk assessment are provided in Table 4.
                           Table 4
          Mass Flux Dilution Factors Used for Chemicals
                In Recreational Risk Assessment
     Chemical
                               Mass  Flux  Dilution  Factor*
                                        (L/M3)
      Acetone
      Aldrln
      Benzene
      Chlorobenzene
      Chloromethane
      Chloroform
      DDT
      DEHP
      1,2-Dlchloroethane
      1,1-Dlchloroethylene
      Dleldrln
      Ethylbenzene
      a-Hexachlorocyclohexane
      Llndane
      Methylene Chloride
      Napthalene
      PCBs  (Arochlor  1248)
      Phenol
      Tetrachloroethylene
      Trlchloroethylene
      Vinyl Chloride
      Xylene
                                       4.8 x 10-5
                                       3.9 x 10-5
                                       2.4 x 10-2
                                       1.4 x 10-2
                                       3.6 x 10-2
                                       1.8 x 10-2
                                       7.6 x 10-5
                                       8.2 x 10-7
                                       3.9 x 10-3
                                       9.5 x 10-2
                                       5.5 x 10-7
                                       3.2 x 10-2
                                       5.0 x 10-3
                                       5.5 x 10-7
                                       8.0 x 10-2
                                       9.9 x 10'4
                                       1.0 x 10-2
                                       1.3 x 10-6
                                       2.2 x 10-1
                                       3.4 x 10-2
                                       7.8
                                       2.0 x 10-2
 *  Assumes exposed person Inhales air driven by five mph
   wind across 100  meters  of contaminated water and mixing
   zone height of two meters.

  The equation used to relate the surface water concentration of
a pollutant to the dose received via inhalation exposure during
swimming and fishing is as follows:
    d = CxMx!xTxfx year/365 days x

                                                \      (4)
                                        70 " 70 I
                              1/bw x  I ฑฃ or ^
where
  C  = concentration of pollutant in surface water (mg/1)
  M  = dilution factor derived from mass flux release algorithm
        (1/m3)
  d   = time-weighted average dose (mg/kg-day)
  I   = inhalation rate (mVhr)
  T  = duration of exposure (hr/event)
  f   = frequency of exposure (events/year)
  bw = average adult male body weight (70 kg)

As with the calculation of dose for dermal risk, the last factor in
                                                                   Equation 4 accounts for the estimation that a person swims only
                                                                   20 years out of a 70-year lifetime or fishes 40 years out of a 70-year
                                                                   lifetime. When exposure parameters from Table 3 were inserted
                                                                   into Equation 4,  the following relationships were obtained:

                                                                   Inhalation dose from swimming =  (1.9  x 10-3)  x C x  M  (5)

                                                                   Inhalation dose from fishing    =  (2.7  x 10 ~3)  x C x  M  (6)

                                                                   Calculation of Dose—Ingestion Exposure
                                                                     Pollutants released to surface waters from NPL sites  may be
                                                                   inadvertently ingested during  swimming. The equation  used to
                                                                   relate the surface water concentration of a pollutant to the dose
                                                                   incurred via ingestion of water during swimming is as follows:
                                                                  d  = G

                                                                  where
                                                                          x  T x  f  x C  x
                          1 liter
                         1000ml
                                                                                                    x?ฐ
                                                                                                    X 70
  year
365 days
   d  = time-weighted average dose of pollutant (mg/kg-day)
   G  = rate of ingestion of surface water (ml/hr)
   T  = duration of each recreational event (hr/event)
   f  = annual frequency of recreational events (events/year)
   C  = concentration of pollutant in surface water (mg/1)
   bw = body weight (70 kg).

 When exposure parameters from Table 3 were inserted into Equa-
 tion 7,  the following relationship was obtained:
 Ingestion dose from swimming =  (3.6 x  10~5)  x C
                                                                                                                          (8)
No significant exposure via ingestion of surface water is expected
during fishing.

Calculation of Excess Lifetime Cancer Risks
  In assessing the excess cancer risk associated with recreational
exposures, the unit cancer risk (UCR) was used as the practical
measure of degree of potential hazard. The unit cancer risk is the
probability of developing cancer if a person is exposed to a pollu-
tant at a dose of 1 mg/kg of body weight/day continuously over
a 70-year lifetime. Unit risk values used in this assessment are shown
in Table 5. Most of the cancer unit risk values were obtained from
the U.S.  EPA's Carcinogen Assessment Group (CAG)12; unit
risks for chemicals lacking CAG derived values (e.g.,  for DEHP)
were obtained from other U.S. EPA sources.13 Due to differences
inherent in the dose-calculation methodologies, inhalation and in-
gestion dose calculations were based on external dose  and dermal
dose calculations were based on internal dose; therefore, separate
unit risks derived from the CAG values were used to calculate risk
due to dermal exposure, as shown  in Table 5.
  The excess lifetime cancer  risk was calculated based  on the
following equation:
                                                                  R  = d x UCR

                                                                  where
                                                                                                                          (9)
                                                                  R  = excess risk or lifetime probability of developing cancer
                                                                  d = the time-weighted average dose (mg/kg-day)
                                                                  UCR = unit  cancer risk (mg/kg-day)
                                                                    By substituting the appropriate expression for dose (i.e., Equa-
                                                                  tion 2 or 3 for dermal exposure, Equation  5 or 6 for inhalation
                                                                  exposure or Equation 8 for ingestion exposure) into Equation 9,
                                                                  the excess lifetime cancer risk was calculated for each chemical on
                                                                  an exposure pathway-specific basis. These risks were then summed
                                                                  across all relevant exposure pathways and chemicals to determine
                                                                  an activity-specific (i.e., relating to fishing or swimming) cancer
                                                                  risk. The cancer risk for the maximum exposed individual (MEI),
                                                                  which is a hypothetical person who engages in both swimming and
                                                                                             HEALTH ASSESSMENT    145

-------
                            Table 5
         Chemicals and Unit Risk Factors for Risk Assessment
                    Of Recreational Exposures.
                                                •Unit Risks
                                       •|nhalซllon/lngซ*llon  "D*rmil
 Pollutant
                           Effect
Acetone
Aldrln
Barium
Benzene
Beryllium
Chlorobenzena
Chloroform
Chloromethane
Chromium
DEHP
DDT
1 ,2-Dlchloroปthanซ
1,1-Dlchloroethylene
(Vlnylldene Chloride)
Dleldrln
Ethylbenzene

a-Hexachlorocyelohexane
Llndane
Mercury
Methylene Chloride
Napthalene
Nickel +*
PCBs (Arochlor 1248)
Phenol
Tetrachloroethylene
Trlchloroethylene
Vinyl Chloride
Xylene
Gl Tract Effecta
Cancer
Trachea
Cancer
Cancer
Kidney/Liver
Cancer
Kidney
Cancer
Cancer
Cancer
Cancer
Cancer

Cancer
Kidney/
Uver Ettecta
Cancer
Cancer
Kidney
Cancer
Ocular
Reproductive
Cancer
Kidney Effecta
Cancer
Cancer
Cancer
Liver Effecta
8.0 x 10-ซ
11.4
3.0 * 10-'
2.9 x 10-i
2.6
2.4 x 10-ป
7.0 x 10-ป
9.85 x 10-*
41.
6.7 x 10-*
3.4 x 10-'
9.1 x 10-*
1.16

30.4
5.8 x 10-*

11.2
1.33
6.9 x 10-1
1.4 x 10->
1.0 x 10-ซ
9.0
4.34
1.0 x 10-*
5.1 x 10-*
1.1 x 10-*
1.75 x 10-*
2.2 x 10-ป
ป
12.0

4.8~TlO-1
	
___
	
	




	

32.0
1.3 X10-*


1.40
	
	

	

1.1 x 10--

	

	
 *  Sources -  References  12 and 13;  units for noncarclnogens are
   (mg/kg-day)-2.
 ** Units for noncarclnogena are (mg/kg-day)*2.
 + Dermal  unit risk factora have been developed only for those chemicals
   that can be assessed for dermal route.
++ While nickel compounds have been Implicated In cancer Incidence In
   some occupational settings (e.g., nickel refineries),  It la unclear
   whether nickel compunds (other than  nickel carbonyl) can cause
   cancer under ambient  exposure conditions.
  fishing, was calculated by summing the risks for the two activities
  for each site.

  Calculation of Risk of Excess Lifetime Noncancer Effects
    The  excess risk of noncancer effects was calculated for each
  pollutant by applying the Weibull model, which is.
  R  = 1-exp [-UNR (d-t)K]

  where
(10)
  R     =  lifetime risk of developing a noncancer effect (unitless)
  d     =  time-weighted average dose (mg/kg-day)
  t     =  threshold dose (mg/kg-day) for a non-cancer effect
  K     =  factor that determines the shape of the dose-response
           curve
           (K = 2 for most non-carcinogens)
  U N R =  unit risk  factor for non-cancer effect
           [(mg/kg-day) -"].

    Accepted unit risk factors for the two selected noncarcinogens
  are shown in  Table 5. Because information on effect thresholds
  generally is lacking, the simplifying assumption was made that the
  threshold dose equals zero; this results in some overestimation of
  risk, but  the accuracy lost for this scoping-level assessment is not
  significant. As with calculation of excess cancer risk, substitution
  of the appropriate dose equation into Equation 10 (setting t  = 0)

 146    HEALTH ASSESSMENT
          allows calculation of chemical-specific excess noncancer risk for
          each exposure pathway. It is technically correct to add these risks
          across chemicals only if the other chemicals involved cause the same
          effect via the same biochemical mechanism.

          Calculation of Site-Specific Risk Ratios
            Site-specific risk ratios are defined as the ratio of the excess life-
          time cancer risk due to ingestion of drinking water to the excess
          lifetime cancer risk due to  recreation, both involving the same
          source of contaminated surface water near a given NPL site. The
          methodology initially involved calculation of the drinking water
          risk  for each  pollutant at each site according to the  following
          equation:
R = C
where
R
C
W
bw =
UCR =
x W x I bw x UCR
excess lifetime cancer risk from consumption
drinking water (unitless)
surface water concentration of pollutant near
(mg/l)
ingestion rale of drinking water (2 I/day)
average adult body weight (70 kg)
unit cancer risk [(mg/kg-day)- ']
(ID
of
the site
Next, these chemical-specific risks were summed for all chemicals
occurring at each site to obtain the site-specific total excess cancer
risk from consumption of drinking water. To obtain the risk ratio
for each site, the drinking water cancer risk was divided by the
recreational MEI cancer risk for the given site.

Calculation of Health-Based Benchmarks for
Recreational Exposure
  Health-based  benchmarks  for  recreational exposure  that
represent surface water concentrations corresponding to 10 '6 risk
were calculated on  a chemical-specific basis. This was done by
solving the appropriate risk equations for concentration for each
exposure  pathway   (swimming-dermal,  inhalation,  ingestion;
fishing-dermal, inhalation) and inserting a risk value of 1  x 10"6
along with the appropriate chemical-specific  input  parameter
values. The minimum of the five pathway, specific concentrations
was then chosen as  the health-based benchmark for exposure to
that chemical via recreational  use of surface water.

RESULTS
Results of Risk Calculations
  The results of the risk assessment are summarized in Table 6.
In assessing these risk numbers, exceedance of a risk of 1  x 10"*
(1 excess lifetime case of given adverse health effect per 1  million
persons)  was viewed as an indication of unacceptable risk.  The
target  risk cutoff used by  some federal agencies to determine
whether specific risks imposed on the public are unacceptable varies
among agencies and program offices. For example, the Food and
Drug Administration typically has used a lifetime risk of 10" * as
the cutoff, whereas  the ambient water quality program at the U.S.
EPA  used a  risk "band" of  10"5  to 10"7 as  the cutoff above
which imposed risks are unacceptable.14
  In this  analysis, no single pollutant generally dominates the con-
tributions to risk for all exposure pathways for any given site. For
example, for Site D, 1,1-dichloroethylene is the major contribu-
tor  to cancer  risk  from  ingestion  exposure during swimming,
whereas vinyl chloride makes the principal contribution to cancer
risk from inhalation  exposure during swimming. This result  demon-
strates that multipollutant approaches, such as that used in this
risk assessment, are needed  to address risk studies at NPL sites.
Although only a few chemicals have  been assessed for non-
carcinogenic health effects,  the results indicate that, in general,
these risks are far below those posed by carcinogenic compounds

-------
                           Table 6
            Summary of Excess Lifetime Cancer Risks
            Associated with Exposure to Surface Water
             Concentrations of Pollutants Near Actual
                National Priority List (NPL) Sites.
                   Excess Lltetlma Cancer Risk

 Site      Swimming     Fishing    Maximum Exposed Individual*

A
B
C
D
E
F"
G
H
1
J
K
L
5.9 X 10-6
3.1 X 10-8
3.6 X 10-6
1.5 X 10-ซ
9.6 X 10-8
7.7 X 10-9
2.8 X 10-5
1.2 X 10-*
1.5 x 10-8
1.4 x 10-7
3.2 x 10-7
7.1 x 10-7
3.2 x 10-8
2.1 x 10-6
1.9 x 10-ซ
1.0 x 10-7
4.8 x 10'13
***
1.7 x 10-*
1.4 x 10-8
7.1 x 10-8
1.6 X 10-8
6.6 x 10-6
6.3 x 10'8
5.7 x 10-6
3.4 x 10-*
2.0 x 10'7
7.7 x 10-ป
2.8 X 10-5
2.9 x 10'4
2.9 x 10-8
2.1 x 10-7
3.4 x 10-7
      Average MEI Risk = 6.1 x 10-5
      *  The Maximum Exposed Individual (MEI) Is a person who
         engages In both fishing and swimming.
      ••  No carcinogens present at site;  therefore not  assessed.
      "• Because the  only carcinogen known to be present at this
         site Is chromium, no appreciable Inhalation exposure
         (due  to low volatility) or dermal  exposure (due to low
         dermal permeability) Is  expected.

and are, in fact, below policy risk cutoffs.
  The excess lifetime MEI cancer risks associated with sites at which
carcinogens are present range across five  orders of magnitude from
7.7 x 10~9 to 3.4 x 10~4. The average MEI excess cancer risk
is 6.1 x  10~5 (median =  3.4  x 10~7).  The site with the highest
risk (Site D) is associated with an excess MEI cancer risk that is
340 times higher than the 10 ~6 risk cutoff and 34 times present
higher than the 10 ~5 risk cutoff. Five of the 11 sites with carcino-
gens (45%) have MEI cancer risks that exceed the 10 "6 risk cutoff;
3 of the 11 sites (27%) have MEI cancer  risks that exceed the 10~5
risk cutoff. Preliminary calculations not included in this  paper
indicate that consideration of consumption of recreationally caught
fish near NPL sites can result in recreational risks that are one to
four orders of magnitude higher than the above risk numbers.
  The percent contributions  of  various  exposure  routes and
activities to total MEI cancer risk were calculated. The major con-
tributors to excess cancer  risk during swimming appear to be in-
halation of volatile organic compounds at five of the 11 sites (45%)
and water ingestion at 5 of the 11 sites (45%), followed by dermal
exposure (9%). Excess cancer risks resulting from inhalation of
volatile organics  during fishing exceed the risks from dermal
exposure for all but one of the sites for which there are  data.
  The contribution of swimming to total MEI cancer risk ranges
from 50 to 100%, with an average of 68% (median  =  63%).
Fishing contributes from 0 to 60% of the total MEI cancer risk,
with an average of 32% (median = 37%). In summary, the risks
of cancer from exposure to Superfund pollutants apparently are
significant at some NPL sites.

Results of Site-Specific Risk Ratio Calculations
  The results of the site-specific risk ratio calculations are indicated
in Table 7. The risk ratios range from approximately 5 to 790 for
the 11 example sites associated with the presence of carcinogens
in surface water. The median value for the risk ratio is 150. While
these results suggest that drinking water  risk may, on the average,
be two orders of magnitude higher than recreational risk from
exposure to contaminated surface waters near NPL sites, the con-
tribution from the recreational component can still be significant
 for specific sites. The latter is particularly true for sites for which
 recreational use of nearby surface waters is expected to occur, but
 for which drinking water use is  not expected.
                            Table 7
              Site-Specific Ratios of Drinking Water
             Cancer Risk to Recreational Cancer Risk.
Site
A
B
C
D
E
F**
G
H
|
J
K
L
Median
Excess Lifetime
Cancer Risk
Drinking Water Recreational MEI*
3.9 x 10-3
7.3 x 10-6
1.5 x 10-4
4.4 x 10-3
1.9 x ID'5

6.1 x 10-6
2.2 x 10-2
1.4 x 10-3
4.3 x 10-6
7.4 x 10*5
2.5 x 10'4
1.5 x 10'4
6.6 x 10'6
6.3 x 10-8
5.7 x 10-6
3.4 x 10'4
2.0 x 10-7

7.7 x 10-9
2.8 x ID'5
2.9 X 10'4
2.9 x 10-8
2.1 x 10'7
3.4 x 10-7
3.4 X 10-7
Risk
Ratio
590
120
26
13
95

790
790
4.8
150
350
740
150
   The 1.0 x 10-6 "target" risk is exceeded at  five (45%)
   of the eleven sites containing carcinogens; the upper
   limit target  risk of 1.0 x 10'5  Is exceeded at three  (27%)
   of the eleven sites.

   No carcinogens at this site; therefore site was not assessed.
Results of Calculation of Health-Based Benchmarks
For Recreational Exposures
  Table 8 presents the calculated surface water concentrations
corresponding to 10 ~6 risk arranged by chemical from lowest to
highest concentration. These numbers represent the minimum con-
centration for each of the five exposure pathways for each chemi-
cal. These benchmarks  do not represent U.S.  EPA-approved
guidelines or standards, but they do provide some rough measure
of "acceptable" versus "unacceptable" risk levels of these pollu-
tants with regard to recreational use. As is evident in Table 8, the
ranges of these minimum concentrations do not substantially over-
lap for cancer versus noncancer effect; these ranges span about
four orders of magnitude for each of these effect categories. The
benchmark concentrations range from approximately 7  x 10~4
to  1 mg/1 for carcinogens and  9 to 1  x  104 mg/1 for non-
carcinogens (the exception is the carcinogen DEHP that has a value
of about 4 x 101 mg/1). These results are reasonable and consis-
tent with the generally lower contribution to risk observed for non-
carcinogens relative to that for carcinogens in the site-specific risk
assessment (i.e., the lower the contribution to risk at a given con-
centration level, the higher is the allowable minimum concentra-
tion corresponding to the 10~6 risk level). There does not appear
to be any chemical class-related pattern to the ranking of these
values, although we have not statistically tested of this hypothesis.

UNCERTAINTIES AND LIMITATIONS
  Other than the usual uncertainties associated with the use of lexicological
data for human risk assessment, there are a number of uncertainties  and
limitations that exist in this risk analysis. These include:

• The  general lack of  site-specific data  on  the frequency  and
  duration of recreational use
• The  availability  of dermal permeability constants for only a  few
                                                                                                HEALTH ASSESSMENT     147

-------
                           Table 8
         Calculated Health-Based Benchmarks for Recreation



Chemical
Chromium
Dltldrln
Aldrln
a-Hexachlorocyclohsxsne
Vinyl Chloride
1,1-Dlchloroethylene
Benzene
PCBS
Beryllium
Linden*
Totrsehloroethyleno
DDT
Chloroform
1 ,2-Dlchloroethsne
Methylene Chlorld*
Trlchloroethylene
Phenol
Nlck.l
Utrcury
DEHP
Barium
Chloromethene
Xylono
Ethylbenzene
Chlorobenzeno
Napthalane
Acetone
* Represents minimum of live
dermel, Ingesllon; fishing •


Terget Orgen
Effect/
Cenoer
Cenoer
Caneer
Cencer
Cancer
Cancer
Caneer
Caneer
Cancer
Caneer
Cancer
Cancer
Caneer
Cancer
Cencer
Cencer
Kidney
Reproductive
Kidney
Ceneer
Trachea
Kidney
Liver
Kidney/Liver
Kidney/Liver
Oculer
Ql Tract
•Surface Water
Concentration
Corresponding to
10-* Risk (mgfl)
I.IE-04
9.JE-04
2.5E-03
2.SE-OJ
2.
-------
                               Environmental  Modeling and the
                        Superfund  Exposure  Assessment  Process

                                                     Peter Tong
                                            Toxics Integration Branch
                                 Office of Emergency and Remedial Response
                                     U.S. Environmental Protection Agency
                                                 Washington, D.C.
                                                   H. Lee Schultz
                                                     Versar Inc.
                                                   Springfield, VA
                                               Seong Hwang, Ph.D.
                                           Exposure Assessment Group
                                      Office of Research and Development
                                     U.S. Environmental Protection Agency
                                                 Washington, D.C.
ABSTRACT
  This paper describes application of environmental modeling tech-
niques to the process used in the assessment of human exposure
associated with uncontrolled hazardous waste sites and provides
an overview of the analytical protocol detailed in the recently
developed Superfund Exposure Assessment Manual. Models are
available to assess contaminant fate and transport in air, surface
water and  groundwater  in terms of analytical and  numeric
solutions.
  Due to variations in resource requirements for a modeling study,
the analyst should consider  whether an exposure assessment can
be accomplished simply by monitoring, by modeling, or by a com-
bination of both techniques and whether the modeling results would
provide enough accuracy for screening/final selection of remedial
action alternatives. Despite the unique characteristics of each site,
the selection and use of modeling can be successfully accomplished
by defining the remedial action criteria, the pathways of contami-
nation, the potential receptors, the data input requirements, the
underlying assumptions used, the inherent uncertainties and the
effectiveness of the model selected.

INTRODUCTION
  CERCLA as amended in  late 1986 by SARA requires, among
other things, that remedial actions conducted  on-site shall meet
the "applicable or relevant and appropriate standards, limitations,
criteria and requirements" (ARARs) of state and federal environ-
mental laws; that remedial actions utilize permanent solutions and
alternative treatment or resource recovery technologies; and that
a health  assessment be conducted on every existing and proposed
site on the NPL. One effect  of these provisions is that U.S. EPA
needs to collect sufficient  data during the RI/FS phase to charac-
terize the risk of the potentially or currently exposed populations
in sufficient detail for risk management decisions. The ultimate
goal is to select a cost-effective remedial alternative that will satisfy
ARARs  and provide adequate public health and environmental
protection.
  Exposure assessment provides a formalized protocol for the
development of data necessary to support such selection. Both
monitoring and modeling techniques are necessary in the develop-
ment  of an integrated exposure assessment.  Indeed, modeling,
whether it is through the use of simple analytical equations or more
sophisticated computer assisted numerical codes, is beginning to
be used increasingly to provide guidance to achieve the above goals.
MODELING VS. MONITORING
  Due to the high degree of complexity in the natural processes,
long, term monitoring and field determinations can help circum-
vent many of the technical difficulties associated with modeling.
Modeling is the scientific attempt to simulate the effects of natural
phenomena. Mathematical models rely on the quantification of
relationships between specific parameters and variables to simulate
the effects of natural processes, e.g., hydraulic conductivity and
contaminant concentrations and rates of movement. In modeling
contaminant transport from an uncontrolled hazardous waste
within a medium or across media, integration of geological, hydro-
logical,  atmospherical, chemical  and biological  disciplines  is
necessary.

  Theories are constantly being developed in an attempt to decrease
or eliminate  uncertainties regarding the true  relationships of
parameters and variables. In reality, however, any models selected
are a compromise between scientific theories, input data acquisi-
tion and economics. Significant error may result  from the dis-
crepancy between the problem conceptualized and the theoretical
approaches (model) taken to tackle that problem. Also, simplifi-
cations of theoretical expressions used to solve practical problems
can cause substantial errors in the most careful analyses5. There-
fore, analysts should never be overly confident of current abilities
to predict contaminant transport and fate.
  The acquisition of high-quality monitoring data is critical to the
successful conduct of an integrated exposure assessment. Initial
monitoring of the subject site provides data necessary to determine
the extent and type of existing on-site contamination. In addition,
off-site sampling can provide direct evidence of the extent of off-
site contaminant migration and of contaminant concentrations at
points of human exposure. From such data, estimates of existing
population exposure can be derived with confidence.
  The exposure assessment process, however, requires more than
a determination of current exposure. When a given site is evaluated,
contaminants  may not have reached points of  potential human
contact. Thus, monitoring may not provide data adequate to
determine  short-term maximum  exposure  levels. In addition,
exposure assessment also requires projection of long-term exposure
(over a 70-year average lifetime). Such exposures cannot be directly
measured in a practical manner. Instead, modeling techniques must
be applied, based on monitored data, to develop  requisite long-
term exposure projections.
                                                                                        HEALTH ASSESSMENT    149

-------
MODEL SELECTION CRITERIA
  Hwang3 delineates three factors which dictate the level of com-
plexity of mathematic models chosen in the selection process:

• Objective criteria— these criteria  refer to the level of modeling
  detail required  to meet the objectives of the study.

  - A screening study is one where limited calibration and vali-
  dation data are  available and the uncertainty associated with the
  predicated results is comparatively large, somewhere  in the
  nature of an order of magnitude. The purpose is to make a
  preliminary screening of a site or  to make a general comparison
  between several sites.
  - An in-depth  study is one where a smaller uncertainty in the
    predicted results is necessary, on the order of a factor of two
    to ten and where the objective is to make an assessment of
    the environmental impact, performance or exposure to recep-
    tors from a specific site. Calibration and validation data are
    necessary in  order to reduce the uncertainty inherent in the
    results and to quantify the margin of error.

 • Technical criteria— these criteria refer to the model's ability to
  simulate site-specific transport and fate.  They are based on the
  physical, chemical and biological characteristics of the site and
  the contaminant of interest. Three categories of technical criteria
  involve:
    Transport  and transformation processes: these criteria relate
    to those significant phenomena  (adsorption, attenuation,
    diffusion,  dispersion, volatilization, complexation, hydroly-
    sis,  oxidation-reduction and biological transformation and
    degradation) known to occur  on-site that must be modeled
    in order to properly represent the site.
  - Domain configuration: these criteria relate to the ability of
    the model  to accurately represent the geohydrologic or the
    atmospheric  system. When high  levels  of  resolution are
    required to predict contaminant concentrations for comparison
    to ARARs, it is generally necessary to simulate site-specific
    geometry and dimensionality  for which numeric (computer
    assisted) solutions are most appropriate.  Use of a simpler
    analytical model may be justified if simplifying the site geome-
    try is geotechnically sound (e.g., water table/confined flow
    system; porous media/fracture flow, steady, state/transient
    flow; single-phase/multiphase flow; constant flow/no flow;
    and one-,  two- or three-dimensional analysis).
  - Media properties: these criteria correspond to the ability of
    the mathematical model to represent the spatial variability of
    site conditions (e.g., are conditions such as hydraulic conduc-
    tivity, recharge and porosity homogeneous or heterogeneous).
 • Implementation criteria: these criteria are dependent on the ease
  with which a  model can be obtained (many models are proprie-
  tary) and have its acceptability demonstrated. The criteria
  identify documentation, verification, validation requirements and
  ease of use so  that the model selected provides accurate and
  meaningful results.

  In an attempt to simulate natural phenomena, most mathematical
 models employ a  few well known equations that relate processes
 with variables under specified assumptions and conditions. Solu-
 tions  to  process equations can range from simple to complex.
 Complexity increases with increase in  accuracy of predictions,
 dimensionality, irregularity of site geometry, heterogenicity of the
 medium, non-uniformity of the flow field (e.g., saturation zone
 vs. unsaturated zone) and disparity in density between the pollu-
 tant and the medium. Therefore, solutions are categorized as ana-
 lytical (simple, mathematical  treatment  involves  the  use of
 hand-held calculators) and numeric (complex, solution to equa-
 tion(s) requires the use of computer codes). Semi-analytical solu-
 tions are ones which fall in between these two  extremes.
  To evaluate air  quality impacts, relatively  simple estimation tech-
niques or analytical solutions that provide conservative estimates
are available for screening studies. The purpose of such techniques
is to eliminate the need of further, more detailed modeling which
requires precise meteorological data inputs. The latter modeling
procedures, based primarily on Gaussian dispersion and  fluid
dynamics, involve a large number of more complex mathematical
computations.  Models of this type are  computer coded12. See
Table  1 for a list of atmospheric fate models.
                           Tibk 1
              Features of Atmospheric File Models
  For analysis of surface water and groundwater contamination,
analytical methods  that handle solute transport in uniform or
porous media are relatively easy to use. To avoid extensive compu-
tations, the analytical solutions available generally are restricted
to time-dependent solute transport in media having steady and
uniform flow. See Table 2 for a list of analytical methods which
can be used to evaluate various remedial actions. Some one- and
two-dimensional solutions are described by Javandel, et  at.4.
Semi-analytical methods  are more  powerful than  analytical
methods, applying the Darcy Law  of fluid mechanics,  the com-
plex velocity potential and numerical tools and computer plotting
                           Table 2
              Remedial Actions versus Simplified and
                   Analytical Methods Matrix
                                  :   !  :i   :•    :
                                  I!  !  Hi  JM   ii   S! I!
                                  j.  ;  j:i  ;;,   :i   ;.- j;
                                                  •    •


                                                  *    *
 150    HEALTH ASSESSMENT

-------
capabilities. The major limitations of semi-analytical methods are:

•  They are only applicable to steady-state and two-dimensional
   flow through homogeneous media
•  The effects of transportation by dispersion and diffusion are
   assumed to be uniform throughout the region of concern

   On the other hand, numerical methods, with the help of a com-
puter, can solve complex equations describing coupled or uncoupled
processes in heterogeneous and  anisotropic formations under
various initial and boundary conditions. Models employing numeri-
cal methods can be categorized as follows according to the approxi-
mation of the first-  and second-order spatial derivative  terms:

•  Finite  difference (FD) method
•  Integrated finite difference (IFD) method
•  Finite  element (FE) method
•  Method of characteristics (MOC)

THE SUPERFUND EXPOSURE ASSESSMENT PROCESS
   The U.S. EPA Office of Emergency and  Remedial Response
(OERR) recently has finalized drafting a Superfund Exposure As-
sessment Manual*.  In this document, the structure and application
of the exposure assessment process are presented along with descrip-
tions of analyses where modeling may be appropriate. The docu-
ment is developed to assist remedial project managers (RPMs), state
officials  and contractors to understand the exposure assessment
process and the various analytical tools available  to  support
component parts of  that process. Criteria and factors for using
certain fate and transport models for exposure analyses are iden-
tified, along with the uncertainties associated with such analyses.
Exposure assessment integrates the evaluation of multiple toxics,
multiple on-site contaminant release sources and multiple exposure
scenarios. It, thereby, provides  a clear perspective regarding the
relative magnitude of each exposure pathway.
   The overall framework for conducting an  integrated exposure
analysis  is illustrated in Figure 1.  It comprises four analytical
segments:

•  Contaminant release analysis
•  Environmental fate analysis
•  Exposed populations analysis
•  Calculation of dose incurred


	
_J,.ซ,.,...,...\

-



lซ'nlt
i



_ 	 _ 	 	 	

                           Figure 1
         Overview of Integrated Exposure Assessment Process
  The results of each segment  serve as input to and provide
direction for the subsequent segment. The process follows the U.S.
EPA  Guidelines for Estimating Exposures13 and addresses the
major areas of exposure assessment advocated by Preuss, et al.6.
ASSESSMENT PROCEDURES
  Specific procedures used in protocol to accomplish these goals
are:

Contaminant Release Analysis
  Contaminant release analysis involves identifying each on-site
source of release of each target chemical to specific environmen-
tal  media. Emissions are characterized  as to types and amounts
of chemicals involved and a determination is made of the level of
release (mass loading) of each chemical to each affected medium.
The results from the release analysis step provide the basis to evalu-
ate the potential for contaminant transport or transformation and
environmental fate. The technique includes screening (qualitative
assessment) and quantitative analyses which predict short-term and
long-term releases for affected environmental media (soil, surface
water, groundwater and air).

Environmental Fate Analysis
  Qualitative determination or quantitative release rates derived
from the preceding analysis provide the basis for this analysis. For
each released  contaminant in each receiving medium, various
environmental transport, facilitated transport, transformation and
removal mechanisms are considered or  quantified. This analysis
is again chemical- and medium, specific and provides a delinea-
tion of the areal extent and magnitude of environmental contami-
nation.  The results of the fate assessment subsequently support
identification of population exposed to contaminants in the ambient
environment  and assessment of exposure levels. The  analysis
includes both qualitative screening of environmental fate pathways
and quantitative analyses, which can involve simplified algorithms
or computer, based models.

Analysis of Exposed Populations
  In this assessment, environmental contamination data from the
preceding two analytical process above are compared with popu-
lations data to determine the likelihood of human contact with
contaminants. Exposed populations analysis begins with a screening
assessment which identifies whether or not the pathways result in
known or potential human contact. Quantitative analysis is con-
ducted by quantifying the populations exposed, characterizing the
high risk groups within the populations and defining the range of
activities of the populations which lead to exposure and the level
of such exposure.

Exposure Calculation and Integration
  Integrated exposure analysis  is conducted for those contaminants
that are released and transported from  the site and that contact
receptor human populations directly or indirectly. Exposure calcu-
lation considers how often receptors come into contact with con-
taminants in specific  environmental media,  the mode  of such
contact and the amount  of contaminant that contacts skin, lungs
or the gastrointestinal tract. The goal is to quantify the amount
of contaminant contacted within a given  time interval or per event
and to project cumulative exposure over a lifetime.

Benefits of Analysis
  The analytical methodology presented in the Superfund Exposure
Assessment Manual provides the user with three distinct benefits.
First, it is by design intended to be applied as a flexible frame-
work to guide site analyses. In some cases, a purely qualitative
analysis may suffice to obtain  requisite insight into the likely level
of risk associated with existing  or projected site conditions. In most
situations, however, some  mix of qualitative  and quantitative
analysis will be appropriate.  Second, the analytical procedures
presented in the manual reflect the current state-of-the-art in con-
taminant release, environmental fate and human exposure assess-
ment techniques. Third, the analytical framework presented in the
"Superfund Exposure Assessment Manual" can be directly applied
to analysis of natural  resource impacts or biotic  population
exposure, although in such evaluations certain analytical methods
                                                                                             HEALTH ASSESSMENT    151

-------
will of necessity vary from those prescribed for human popula-
tion exposure assessment.

UNCERTAINTY
  The uncertainty involved in the use of mathematical models in
exposure assessment cannot be overemphasized. Uncertainties are
due to model design as a simplified representation of reality and
to the variability  or accuracy of the input parameters used in
modeling. The former limitation can be characterized by exten-
sive documentation of detailed mathematical testing to determine
the magnitude of errors generated by the assumptions and simpli-
fications involved5.  For example,  the assumption of  no-
interaction among the solvents is partially reputed by the  identi-
fication of facilitated transport of hydropholic organic chemicals
in mixed solvent systems2-l7 -u.
   Serious mass balance errors due  to assumption of linearity of
the sorption term have resulted in underestimation of contaminant
mass in the solution phase and overestimation of contaminant mass
in retardation by sorption to soil particles. Similarly, due to the
lack  of appropriate  field  determinations of  natural process
parameters (not adequately addressed by sensitivity or stochastic
analyses), the release or transport and fate of contaminants being
modeled also may be erroneous. Hydraulic conductivity,  for in-
stance, may vary greatly because of differences in density and vis-
cosity of fluids that are present. The result is a dramatic downshift
in local flow directions near plumes that have as little as  1 % in-
crease in density relative to uncontaminated water5. Thus, field
verification by calibrating the models used and benchmarking rou-
tines (comparing the efficiency of different models in solving the
same problem) plus adherence to stringent model selection criter-
ia are perhaps the best approach to diminish uncertainty.

CONCLUSION
   Characterizing risk posed by uncontrolled hazardous waste sites
necessitates a thorough evaluation of current and future exposure
and the hazard of the pollutants involved. The risk estimated there-
fore represents the risk posed by a "no action alternative." Even
when remedial actions are taken, there may  be some risk  of
exposure depending on the degree of protectiveness and long-term
reliability of each alternative. Hence,  the exposure assessment
performed should provide an objective representation of:  (1) the
seriousness of the  problem, (2) the potential benefits obtained by
various remedial  alternatives. The analyses should  be able to
provide sufficient inputs for a risk management decision, i.e., final
selection of a remedial alternative. This paper describes the poten-
tial role of mathematical modeling in exposure analyses and the
procedures used in an objective assessment of the problem posed
by an uncontrolled hazardous waste site.

REFERENCES

 I. Bonazountas M., et al. "Environmental mathematical pollutant fate
   modeling handbook/catalogue" (Draft): U.S. EPA, Office of Policy
   and Resource Management, Washington, DC, 1982
 2. Enfield  C. O. "Chemical transport facilitated by multiphase flow
   systems," Seminar on degradation, retention and dispersion of pollu-
   tants in groundwater, Copenhagen,  Denmark, Sept.  1984
 3. Hwang, S. T. "Model selection  criteria for performing exposure as-
   sessments: groundwater models." Office of Health and Environmen-
   tal Assessment. U.S. EPA, Washington, DC, 1987
 4. Javendel L,  Doughty C. and Tsang C.F. Groundwater Transport:
   Handbook of Mathematical Models. American Geophysical Union
   Water Resources Monograph Series 10. Washington. DC, 1984
 5. Keely J. F. "The use of models  in managing groundwater protection
   programs."  U.S. EPA. Robert S. Kerr Environmental Research
   Laboratory.  Ada. O.K. 74820.  EPA/600/8-87/003,  1987
 6. Preuss P.W.. Ehrlich A. M. and Oarrahan K. G. "U.S. EPA guide-
   lines for risk assessment." /Vac. of the 7th National Conference on
   Management of Uncontrolled Hazardous Waste Sites. Washington,
   DC, 1986
 7. Rao P.  S. C. Hornsby A.  G., Kilcrease D. P. and Nkedi-Kizza P.
   "Sorption and transport of hydrophobic organic chemicals in aqueous
   and mixed solvent systems: model development and preliminary evalu-
   ation."  J. Environ. Qual. 14. No.  3. I98S
 8. Schulu  H. L.. Palmer W. A., Dixoo G. H.. Gteit  A.  F..  Mercer
   M. L. and Dickenson R. A. Superfundexposure assessment manual.
   Draft Report. Office of Emergency and Remedial Response, VS. EPA,
   Washington, DC, 1987
 9. U.S. EPA. Environmental modeling catalogue. U.S. EPA, Washing-
   ton, DC. 197S
10. U.S. EPA. Environmental modeling catalogue. U.S. EPA, Washing-
   ton, DC, Information Clearing House. PM-21IA.. 1982
11. U.S. EPA. Modeling remedial actions at uncontrolled hazardous waste
   sites. U.S. EPA. Washington, DC.  I98S
12. U.S. EPA. "Guideline for air quality models." Office of Air Quality
   Planning and Standards, U.S. EPA. Research Triangle  Park. NC. 1986
13. U.S. EPA. "Guidelines for exposure assessment." Office of Research
   and Development, U.S. EPA, Washington, DC. 51 PR 34042., 1986
14. Woodburn K. B.. RAO PSC, Fukui  M., Nkedi-Kizza P. "Sph-opho-
   bic approach for predicting sorption of hydrophobic  organic chemi-
   cals on synthetic sorbenls and soils." J. of Contain. Hyd. 1:1986,
   227-241.
152    HEALTH ASSESSMENT

-------
                       An Exposure  Assessment  Modeling  System
                                    For  Hazardous  Waste  Sites
                                                  Alison C. Taylor
                                               Gradient Corporation
                                             Cambridge, Massachusetts
                                            David E. Burmaster, Ph.D.
                                              Burmaster & Associates
                                             Cambridge, Massachusetts
                                              Brian L.  Murphy,  Ph.D.
                                               Gradient Corporation
                                             Cambridge, Massachusetts
                                                  Scott  H.  Boutwell
                                              Independent Consultant
                                              Arlington, Massachusetts
ABSTRACT
Exposure assessment is a critical component of the risk assessment
process. It provides concentration data at receptor points which
can be used to characterize risks with dose-response information.
  Given the many potential sources, transport pathways, receptor/
exposure routes and the relationships between them, there is a need
to organize exposure techniques and algorithms in a comprehen-
sive framework that allows for flexibility of technique selection
according to scenario. To accommodate these needs, the authors
have developed an exposure assessment modeling system comprised
of over 40 "modules" that simulate emission, transport and fate,
and receptor loading processes. The modeling system is embodied
in Lotus l-2-3(TM) spreadsheets which can be run on a microcom-
puter. The modules run on an interactive basis making them easier
to use.  The accompanying workbook also provides the user with
guidance on parameter estimation as well.
  The modules in this system utilize simplified representations of
complex phenomena in order to have analytical (or algebraic) solu-
tions to the governing partial differential equations. Accordingly,
they are used to screen potential exposure scenarios, identify gaps
in the available site data, assess the need for higher resolution
modeling and check the reasonableness of the results of the more
sophisticated models when they are used. Many of the modules
concern peak  concentrations or concentrations along the center-
line of  a plume and represent steady-state or equilibrium condi-
tions. The approach chosen can simulate long-term conditions (e.g.,
seasonal or annual) but not dynamic conditions (e.g., hourly). This
approach is consistent with risk assessment for chronic toxicity,
such as carcinogenicity.
  The transport and fate modules encompass the complete range
of environmental compartments  including: outdoor air, indoor air,
surface water, fish tissue, sediments, soils, house dust, soil gas,
soil moisture (vadose zone) and  groundwater. For exposure path-
ways into the  human body, the system includes modules  for: in-
gestion of contaminants in food, water, soil and dust; inhalation
of contaminants in dust and indoor or outdoor air;  and dermal
absorption from contact with contaminated soils, water or house
dust. The modules also allow the calculation of fluxes within and
among media and receptors,  to  evaluate  complete  exposure
pathways.

INTRODUCTION
  Under the sponsorship of the Gas Research  Institute's (GRI)
Program for Manufactured Gas Plant (MGP) Sites, the authors
have collected a set of simple analytic models that can be used to
estimate exposure concentrations at hazardous waste sites in general
and at MGP sites in particular. In addition, the models can be used
for sensitivity analyses to determine the more important chemical
groups and exposure pathways at a  particular site and their
attendant uncertainties. These models provide the tools for inte-
grated analyses as a part of the overall GRI program for: (1) charac-
terizing former MGP sites, (2) assessing the public health and
environmental risks that such sites may pose with regard to pro-
posed uses and (3) evaluating proposed remedial actions.

OBJECTIVES FOR THE MODELS
  These  models  and  the  associated  spreadsheets  have two
objectives:

• To provide site owners with tools to analytically model the emis-
  sion, transport, and fate processes which determine contaminant
  concentration in the environment resulting from waste disposed
  at sites; these tools represent one facet of the risk assessment
  methodology,
• To facilitate sensitivity analyses with regard to  model input
  assumptions, presence of chemicals in particular media, chemical
  properties and exposure pathways; this task is accomplished by
  documenting the general relationships among sources and trans-
  port pathways  and providing the models in a set of Lotus
  1 -2-3 (TM) spreadsheets.

  These analytic models include ones for phenomena within and
among most of the important environmental compartments. They
can be used to estimate long-term exposure of individuals who come
into contact with contaminated sites. When used in conjunction
with site specific characteristics, the models provide the tools neces-
sary to screen sites to identify those where more study is necessary
and/or where remediation may be necessary. When combined with
information about toxicological effects, the models allow the esti-
mation of chronic human health risks for current conditions at
the  site, conditions sometimes referred to as the "base case" con-
ditions.
  If a site is being considered for new economic uses, the models
will help determine the need, if any, for remedial alternatives. To
a lesser extent, the models also may be used to estimate exposures
of workers and neighbors to chemicals mobilized on and  off the
site during the implementation of remedial measures.

FRAMEWORK FOR THE MODELS
  Figure  1 shows the framework chosen for these models; it can
be described best as a multi-media,  multi-pathway and multi-
process "box" framework of general applicability. More specifi-
cally, the framework  includes compartments for indoor and out-
                                                                                        HEALTH ASSESSMENT     153

-------
                           Figure 1
          Exposure Model Framework with Number Codes
                  Representing Chosen Models

 door air, surface and groundwater, soils, soil gas, sediments and
 fish — each as a separate environmental medium in a separate com-
 partment with fluxes between the compartments.
   The framework specifically tracks the transport and fate of
 chemicals as they move downwind, down stream and down-gradient
 from sources (e.g., wastes above, at or below grade) to receptors
 (e.g., people living or working  on or near the site).  The frame-
 work recognizes that several processes may cause flux of com-
 pounds from one compartment to another (e.g., organic chemicals
 may move from the soil compartment to the air compartment by
 volatilization or by the levitation of fugitive dust to which they
 are adsorbed). The framework also explicitly recognizes the three
 routes into the human body: ingestion of solids (e.g., food and
 contaminated soils) and  liquids  (e.g., water and  beverages);
 inhalation of vapors and fugitive dust; and dermal penetration from
 direct contact.
   Many exposure and risk assessments do not include all of these
 possible processes. For example, a Superfund risk assessment may
 only consider actual or hypothetical exposures (1) via direct inges-
 tion of surface or groundwater,  (2) direct contact with contami-
 nated soil on site and (3) ingestion of fish. These less-than-complete
 analyses  may overlook other  key pathways, particularly air-
 mediated pathways and indoor pathways. Such analyses may con-
 sider hypothetical pathways (e.g., the direct consumption of
 groundwater at sites where no active wells are present), while over-
 looking  actual pathways (e.g., the entrainment of contaminated
 soil gas  into heated buildings).
   The spreadsheet framework encourages holistic analysis. Used
 effectively at the beginning of a project, it can promote the iden-
 tification and analysis of the key exposure  pathways.
   Of course, the framework as presented does not include every
 conceivable pathway  that may exist in nature or that may be
 important in certain circumstances. For example, the current system
 does not explicitly include the models describing the wet or dry
 deposition of air pollutants onto forage grass or food crops which
 could lead to possible human exposure in cow's milk or direct in-
 gestion,  respectively. However, additional algorithms are readily
 accommodated in the spreadsheet format.

 IDENTIFICATION AND SELECTION OF MODELS
   In choosing specific models, the authors  considered a variety
 of potential models from a number of sources. For some of the
 exposure pathways, there was an obvious choice of model based
 on the personal experience of members on the project team. Their
 ability to recall and locate models from past experience was a major
 asset in  assembling such a comprehensive framework without a
sizeable research budget. Other sources of potential model choices
included the U.S. EPA's Superfund Exposure Assessment Manual
(1), and other existing handbooks.
  Several criteria were used to when evaluate and select models
for inclusion in the general framework. A model had to: (1) rest
solidly on first principles from physics, chemistry or biology, (2)
be algebraic or analytic in nature to allow incorporation in a spread-
sheet, (3) apply to long-term exposures appropriate for the esti-
mation of chronic exposure and health risks, and (4) be practical
in  application (i.e., require readily available  parameters).

RESULTS
  Models were drawn from the above mentioned sources for the
pathways of exposure identified in the framework. There were a
few pathways for which no appropriate existing model could be
located. These pathways were identified as areas which will require
further research. For example, we did not locate models describing
pathways involving NAPL (non-aqueous phase liquid) transport
in  a form suitable for coding in a spreadsheet program.
  Figure 1 shows the relationships among the models chosen to
fit the framework. In most instances, no single model suffices for
a particular flux or transport process. Therefore the user has some
flexibility in choosing models to represent certain transport path-
ways. Along with this  flexibility comes a certain  amount of
responsibility. An analyst must know which set of models pertains
to a particular scenario and be able to apply the sequence of models
correctly. For example, an  analyst must use best engineering
judgment to choose one or several  of seven  models  useful for
describing the flux of contaminants from "Wastes at Grade" to
"Air on Site." The different models predict the rate of volatili-
zation of vapors from wet and dry soils and the levitation of fugitive
dust in different situations, including truck traffic and/or excava-
tion of contaminated soil.
  To model complex scenarios, the equations modeling each step
of the pathway are linked by the  analyst (the output from one
becomes the input to the next). By mixing and matching among
the models, some rather complex pathways may be represented.
One such complex pathway is discussed in detail in the example
below.

SPREADSHEETS
  In order to facilitate repeated use of the selected models a Lotus
l-2-3(TM) spreadsheet  format was  develo'ped into which each
model was coded. Figure 2 shows  the generalized representation
         tM In p*riicwlซ<* mi
                                                0 .. i M 100
 •tor of 
-------
of the screen format. Although each model varies in its number.
and type of inputs, intermediate results and necessary constants,
the spreadsheets adhere to a single basic design. The top section
contains the model output, the center section receives the required
inputs and the bottom section contains the equations for calcu-
lating intermediate results of the model.
  There are many advantages to representing the models in this
spreadsheet format. Spreadsheets are interactive and easy to use
and therefore encourage an iterative approach to problem solving.
The fact that intermediate results of a single model are visible to
the user makes the system even more transparent and accessible
than if  only the final output of a model were available.
  The spreadsheets are designed to calculate a final output only
after all of the inputs are entered. The entire spreadsheet can be
viewed but only the input spaces can be filled in or edited. Appro-
priate units of measurement are defined beside each input and
where possible, default values for the inputs are provided. Since
many of the inputs span a large range or rely heavily on site specific
information, this is not  always possible.

EXAMPLE
  As an example, the exposure of an individual living or working
in a house or trailer on-site to the suspension of contaminated par-
ticulate matter from vehicle traffic is modeled. Penetration through
the building shell and subsequent inhalation of contaminated par-
ticulate matter from  the site are estimated. As  in any exposure
modeling problem, the complete pathway must be broken into
individual steps such  that contaminant concentrations in various
compartments—and ultimately dose—can be calculated. The path-
way examined here can be broken down into four basic stages. In
the first stage, contaminated particulate matter leaves  the ground
surface and is suspended in the on-site air due to the action of truck
wheels. The second stage  involves the mixing of the suspended mass
of contaminant in the "box" of on-site air. In the third stage, the
air inside a building on-site reaches a steady-state concentration
of contaminant through indoor-outdoor air exchange. The final
stage is the inhalation of the contaminant by the individual inside
the building.
  There are four calculations necessary to calculate the dose to
the individual inside the building in this problem. The first calcu-
lation determines  the mass flux of contaminated soil from the
ground surface on-site into the on-site  outdoor air due to vehicle
traffic, see Figure  3. Figure 4 contains an analysis of this stage
with  fabricated data to  illustrate the use of the Vehicle Traffic
model. Next,  using the Near Field Box model  shown in Figures
5 and 6 to describe the mixing of the suspended contaminated par-
ticulate in the on-site air, the concentration of contaminant in the
air on-site is calculated. In the third step, the concentration of con-
taminant in the air inside the on-site building is determined based
on an air exchange model. The output of the Near Field Box model
(the concentration of contaminant in the outdoor air) is used as
                                        This model calculates  the mass flux of contaminant from  the
                                        ground soil onsite  to  the air due to vehicle traffic.
    Wind
                 c
w
                              w-
                              X "
                            Figure 3
       Particulate Suspension due to Vehicle Traffic—Diagram
        10
                                                   Vkj
Q

Q   emission rate of particles 10 microns and smaller (mg/hr)

a = mass fraction of contaminant in particulate emissions  (ppm)

V)j   average vehicle-kilometers travelled onsite in one hour,
     totalled across all  vehicles (hr~ )

E1Q   .85(s/10)(S/24)-8(W/7)-3(w/6)1'2((365-p)/365)

EIQ   emission factor for an unpaved road per vehicle-kilometer
      of travel  (kg)

s    percent silt in road surface (Q < s * 100)

S    mean vehicle speed (km/hr)

w    mean vehicle weight  (metric tons)

"    mean number  of  tires (unitless)

p    number of days  per year with at least .254 mm (.01 inches)
     of  precipitation (unitless)

For our  example the  following values will be used:

a         0.1  (ppm)               w         6    (unitless)
                                                  0.1   (ppm)
                                                  8  (%)
                                                  24 (km/hr)
                                                  15 (metric tons)
                                 P
                                 vk
                                            100   (unitless)
                                            24    (hr"1)
                                        The  intermediate result, E,.,  is 0.62  kg.

                                        The  final result, Q, is 1.49 mg/hr.


                                                                    Figure 4
                                                              Vehicle Traffic Model2
                                           Wind
                                           Ground
                                                       W,
                                                                                              m
                                                                    Figure 5
                                                         Near Field Box Mixing—Diagram
the input to this air exchange model, see Figures 7 and 8. Finally,
the dose of contaminant to an individual inside the building is calcu-
lated based on the indoor air concentration of contaminant and
a breathing rate as shown in  Figure 9.
  Each of the models included in the framework incorporates
assumptions as to the conditions it covers. Each of the four models
discussed in the example above carries assumptions. In the Vehicle
Traffic model, uniform vehicle speed, a fixed distance from the
moving vehicle to the receptor and uniform percentage of silt across
travelled soils, are all assumed. The Near Field Box model assumes
uniform wind speed and uniform mixing throughout the box. The
air exchange  model includes two factors which reduce long-term
indoor air concentrations relative to outdoor concentrations. The
penetration factor, p, accounts for the fact that large particles be-
come lodged in the building shell without transport into the interior.
The loss factor,  d, represents the fact that deposition on indoor
surfaces further reduces the suspended indoor concentration. The
                                                                                                HEALTH ASSESSMENT     155

-------
This model calculates  the contaminant concentration in ambient
air onsite after  mixing has occurred.

C0   Q/(HbWbUm)

C0   concentration of  contaminant in ambient air  on site  (mg/m3)

0    emission rate of  contaminant; this input is  the output  from
     the Vehicle  Traffic Model, Figure 4 (mg/s)

Hb = downwind height of box, Hb depends on x, the downwind
     distance which is assumed here to be 100 m (m)

Wb   width of box, crosswind dimension of area of contamination,
     (m)

UTO -  average wind speed through the box, calculated below (m/s)

     u,,,  • 0.22 (U1()) ln(2.5 Hb)

     u.- = wind speed  at 10 m, wind direction ii  not
     critical in  this  model (m/s)

For this example  the following input values are used:


0    =    1.49 mg/hr
Hb        6.2 m         (2)
Wb   =    100 m
u10       4.0 m/s

The intermediate  result, um .  2.41 m/s.

The final result  is Co = 0.001 mg/m3 (or 1 ug/m3),  the
contaminant concentration in the ambient air onsite.
This model calculates the indoor air concentration of contaminan
due to indoor-outdoor air exchange.


Ci - (P)C0 + S/la+d)

ct • concentration of contaminant  in indoor air  (mg/m3)

co • concentration of contaminant  in outdoor air  lmg/m3)

p  * particulate penetration factor (unities*)

     p * f(a/a+d) where f equals the fraction of  particulate
     which becomes lodged in the building  (hell,  and, a and d arc
     as defined.

ซ  • natural ventilation air exchange  rate (hr"ll

d  • removal rate of contaminant inside  (removal  due to
     deposition) (hr'M

S  • Indoor source strength,  this  tern is  included only when
     there is Indoor production  of, or contribution to, the
     contaminant level (mg/B3-hr)


The following input values are used in this calculation:

C.   •    .001  mg/m3     from the  calculation in  Figure 6
p    •    0.4            an estimated  value for p (7)
S    ป    0              assumes there is  no indoor production of
                         the  contaminant
The concentration of  contaminant  in the indoor air is
mg/m3 (or 0.4 ug/m3).
                                                          0.0004
                           Figure 6
                   Near Field Box Model3- 4-5
                             Figure 8
               Indoor-Outdoor Air Exchange Model6- 7
    Wind
                                                                    This model  calculates the dose to an individual breathing air
                                                                    contaminated with concentration Ci at the breathing rate Ir.
                            Figure 7
              Indoor/Outdoor Air Exchange—Diagram


 inhalation dose is calculated assuming 100% retention of the par-
 ticulate pollutant and does not consider the exhaled mass of pollu-
 tant, an assumption appropriate for  use with potency factors
 developed by the U.S. EPA.

 NEXT STEPS
   Putting together this collection of models has been a rich learning
 experience for all of the individuals involved in the project. The
 framework has proved to be robust  and relatively complete in
 solving the exposure modeling problems to which it has been
 applied thus far.
   There are many areas in which this project could be expanded
 or improved. A comprehensive sensitivity analysis on each model
 would help characterize the uncertainty which accompanies results
 from the models. As for any modeling system, it is important for
 an analyst to know whether or not the model outputs will degrade
 gracefully for extremes of input values. Another expansion would
 involve developing models for the exposure pathways for which
 no existing models could be located. In some cases there are existing
 models for main-frame computers which are suitable for future
                                                                    Inhalation  Dose  -
                                                                                         Ir Te
 Inhalation Dose  = the mass of contaminant which an individual
      inhales  (ing/day)

 C,    ป    concentration of contaminant in the air available to
          the individual at the time of exposure (mg/m3)
                                                                              the  inhalation rate of the individual in question
                                                                              (m3/hr)

                                                                              time of exposure, the amount of tine that the
                                                                              individual is exposed to the air containing the
                                                                              concentration C  of contaminant (hr/d)
The following input value* are used for the example:

GI   -    0.0004 mg/m3   from calculation in Figure 8
Ir   =    1.2 mVhr      estimated inhalation rate for an adult
                         during light activity (8)
Te   ซ    8 hr/d         assumes an 8 hour exposure daily


The resulting Inhalation Dose is 0.0038 mg/day.
                            Figure 9
                      Inhalation Dose Model
 adaptation for use on personal computers.
   Some initial work to determine which gap in the framework is
 most limiting in commonly encountered modeling problems would
 be appropriate.  The system's lack of ability to do "vector
 processing" (i.e., perform calculations for multiple compounds
 at a time) has been and will continue to be a drawback in projects
 where the effects of a group of chemicals are being studied. As
 a final improvement, this system of models could be linked to auto-
 matically feed the output of one model to another. However, as
 was mentioned above, the ability to peer into the system at various
 points and verify that the models are responding appropriately to
 a given situation is worth preserving in any integration of the
 modules into an overall systems model.
156    HEALTH ASSESSMENT

-------
REFERENCES
1.  Schultz, H.L., et al., "Draft Superfund Exposure Assessment Manual,"
   Office of Solid Waste and Emergency Response, U.S. EPA, Washing-
   ton, DC,  1986.
2.  Cowherd, C., O.E. Muleski, P.J. Englehart and D.A. Gillette, "Rapid
   Assessment of Exposure to Paniculate Emissions from Surface Con-
   tamination Sites," Midwest Research Institute, Kansas City, MO, 1984.
3.  Environ  Corporation,  "Potential  Health  Risks  from  Former
   Manufactured-Gas Plant Sites: Exposure Assessment," Washington
   D.C.,  1986, 61-125.
4.  Pasquill, F., "The Dispersion of Material in the Atmospheric Boundary
   Layer—the Basis for Generalization," In: Lectures on Air Pollution
   and Environmental Impact Analysis, American Met. Soc., Boston, MA,
   1975.
5.  Horst, T.W., "Lagrangian Similarity Modeling of Vertical Diffusion
   from a Ground Level Source," Int. Applied Met., 18, 1979, 733-740.

6.  Murphy, B.L., and J.E. Yocom, "Migration Factors for Particulates
   Entering the Indoor Environment," APCA 79th Annual Meeting,
   Minneapolis, MN, June 1986.

7.  Murphy, B.L., and J.E. Yocom, "Final Report on Environment-Indoor
   Migration Factors," TRC Environmental Consultants, Inc., Athens,
   GA, 1986.

8.  Synder, W.S., M.J. Cook, E.S. Nasset, L.R. Karhausen, G.P. Howells,
   and I.H. Tipton, "Report of the Task Group on Reference Man," In-
   ternational Commission on Radiological Protection Papers 23, Perga-
   mon Press, NY,  1975.
                                                                                                    HEALTH ASSESSMENT     157

-------
                     Human  Exposure Potential  Ranking Model

                                                   Lee Ann  Smith
                                             Law Environmental,  Inc.
                                          Government Services  Division
                                                  Atlanta, Georgia
                                                Cynthia  D.  Patrick
                                            ICAIR, Life Systems,  Inc.
                                                  Cleveland, Ohio
                                                Charles M. Hudson
                                     Bureau of  Toxic Substance Assessment
                                     New York  State Department of Health
                                                Albany,  New York
ABSTRACT
  This paper presents an overview of a recently completed project
to finalize the development and implementation of the Human
Exposure Potential Ranking Model (HEPRM) for the New York
State Department of Health (DOH). The HEPRM was developed
to rank hazardous waste sites in terms of the potential for human
exposure to contaminants released from the sites. The  HEPRM
generates a ranking score, based on the probability of human
exposure, for all the hazardous waste sites currently included in
the DOH investigation program. The HEPRM ranking score will
be used to set priorities for further investigative and  remedial
actions at the sites.
  The general ranking scheme is to gather environmental data and
information on a site, to characterize the site as a source of con-
tamination and then to proceed in a logical fashion to assess and
estimate to what extent each of the defined environmental path-
ways could result in  human exposure. The model output (i.e.,
ranking score) for each pathway is calculated as the product of
the Chemical Factor Score, the Probability Score, the Target Factor
Score and the Weighting Factor Score.  The individual pathway
scores for each medium are summed to yield an overall Medium
Score (e.g., Air Score, Surface Water Score, etc.). The Medium
Scores are, in turn, summed to yield an overall Site Score. The
resultant numerical score for any single site has meaning only when
one site is compared to another. By comparing the scores for all
sites, the relative priorities for  the investigation and remediation
of sites is established.

INTRODUCTION
Article 27, Section 1305 of the New York State Environmental Con-
servation (Solid Waste) Law mandates that the New York State
Department of Environmental  Conservation (DEC), in coopera-
tion with the New York State Department of Health (DOH), shall
evaluate existing site evaluation systems and develop a system to
select and prioritize inactive hazardous waste sites for remedial
actions. The law further states that the developed  system shall
incorporate environmental, natural resource  and public health
concerns.
  There are currently three site evaluation systems in use or under
development by DEC to prioritize sites. The DEC Division of Solid
and Hazardous Waste evaluates waste sites in the state in accor-
dance with  the Hazard Ranking System (HRS).  In response to
perceived limitations in  the HRS, DEC and DOH decided to
develop additional site evaluation systems for use by the state in
selecting and prioritizing sites for remedial action. These additional
site evaluation systems are the Human Exposure Potential Ranking
Model (HEPRM), developed by DOH, and the Biothreat Model
which was developed by the DEC Division of Fish and Wildlife.
  The objective of this paper is to present an overview of the
HEPRM1. The HEPRM was developed because the DOH did not
feel that the HRS adequately addressed public health concerns.
The HEPRM is based on the HRS and is therefore capable of
incorporating  HRS input data.  Additionally,  the HEPRM
addresses HRS deficiencies including the estimation of potential
contaminant releases in the absence of environmental data and the
potential for human exposure via all possible migration routes (i.e.,
air, groundwater, surface water and soil).


OVERVIEW OF THE HEPRM
  The HEPRM is designed to fully consider human exposure via
ingestion, inhalation and dermal contact. The model creates a
mechanism to assess all of the significant environmental pathways
that may lead to human exposure. It also provides a mechanism
to estimate the potential for exposure in the absence of data. Table
1  summarizes the human  exposure  pathways addressed by the
HEPRM.
  The HEPRM scoring methodology is illustrated in Figure 1.  As
shown, the model output (i.e., ranking score)  for each exposure
pathway is calculated as the product of the Chemical Factor Score,
Target Factor Score, Probability Score and Weighting Factor Score.
  The individual pathway scores for each medium are summed to
yield an overall Medium Score  (e.g..  Air Score, Surface Water
Score, Groundwater Score, etc.). The Medium Scores are, in turn,
summed to yield an overall Site Score. The resultant numerical score
for any single site has meaning  only when one site  is compared
to another. By  comparing the scores for all sites, the relative pri-
orities for investigation and remediation of sites are established.
COMPONENTS OF THE HEPRM
  As  discussed  above, the four  components of the individual
exposure pathway scores within the HEPRM are the Chemical
Factor Score, Target Factor Score, Probability Score and Weighting
Factor Score. Each exposure pathway is defined by a unique com-
bination of these components. For example, the pathway score for
the ingestion of potentially contaminated groundwater would be
calculated as the product  of  the  Chemical Factor Score  for
groundwater-contact,  the Target Factor Score for groundwater,
the Probability Score for the transport of groundwater off-site and
the Weighting Factor score for the ingestion of water. Each com-
ponent of the pathway score is  discussed in greater detail below.
158    HEALTH ASSESSMENT

-------
AIR
                       Table 1
  Human Exposure Pathways Considered by the HEPRM


Inhalation of air vapor (onsite)
Inhalation of air vapor (offsite)
Inhalation of participates (onsite)
Inhalation of participates (offsite)
Inhalation of soil  vapor  (basement)
SOIL
      Ingestion of soil (onsite)
      Dermal contact  with soil (onsite)
      Ingestion of plants  (onsite)
      Ingestion of airborne soil  (offsite)
      Dermal contact  with airborne soil  (offsite)
      Ingestion of plants contaminated with airborne soil (offsite)
      Ingestion of waterborne soil (offsite)
      Dermal contact  with waterborne soil (offsite)
      Ingestion of plants contaminated with waterborne  soil  (offsite)

CROUNDWATER

General:
      Tngestion of groundwater  (water supply)
      Inhalation  of groundwater vapors (water supply)
      Dermal contact  with groundwater  (water supply)
      Inhalation  of groundwater vapors (basement seepage)
      Dermal contact  with groundwater  (basement seepage)
      Dermal contact  with seepage
      Inhalation  of seepage vapors
      Ingestion of plants  (irrigation)

Surface Water Recharged by  Contaminated Groundwater:
      Ingestion of surface water (water supply)
      Dermal contact  with surface water (water supply)
      Inhalation  of surface water vapors (water supply)
      Ingestion of surface water (recreation)
      Dermal contact  with surface water (recreation)
      Ingestion of plants  (irrigation)
      Ingestion of aquatic biota

SURFACE WATER

General:
      Ingestion of surface water (water supply)
      Dermal contact  with surface water (water supply)
      Ingestion of surface water (recreation)
      Dermal contact  with surface water (recreation)
      Ingestion of plants (irrigation)
      Ingestion of aquatic biota

Surface Water Receiving Runoff from  Lagoon  Overflow:
      Ingestion of surface water (water supply)
      Dermal contact  with surface water (water supply)
      Inhalation  of vapors from  surface water (water  supply)
      Ingestion of surface water (recreation)
      Dermal contact  with surface water (recreation)
Chemical Factor Score
  The Chemical Factor Score is used to select the contaminants
of concern for a site based on their toxicity and physical/chemical
properties.  Figure 2  illustrates the methodology  for selecting
contaminants of concern for a site.
INDIVIDUAL PATHWAY
SCORES
(CF x TF x P x WF)
*
E PATHWAY SCORES =
SITE SCORE
   DEFINE
'HYSICAUCHEMICAL
  PROPERTIES
    AND
   TOXIC1TV
                                                                         DETERMINE
                                                                         MIGRATION
                                                                         POTENTIAL
                                                                CALCULATE GENERAL
                                                                 CHEMICAL FACTOR
                                                                 SCORES FOR EACH
                                                                   CATEGORY

(m

n HJJ)



(ppm)
EVAPORATION
POTENTIAL
EP SCORE - 1-1)
	 ,1, '— ~
1 '('
SOIL-VAPOR
EP x TOXICITY

1 	 i
LEACHING
POTENTIAL
(LP SCORE - 1-3)
GROUNOWATER-CONTAC


CROUNDWATER-VAPOR
HLP * EP111 • TOXICITY



                                                                                                   SELECT TOP FIVE SCORES
                                                                                                    IN EACH CATEGORY FOR
                                                                                                  CONTAMINANTS OF CONCERN
                             Figure 1
                    HEPRM Scoring Methodology
                                                                                            Figure 2
                                                                        Methodology for Selecting Contaminants of Concern
                                                                                 Based on Chemical Factor Scores
                                                                  The first step in the selection process is to generate a list of all
                                                                chemicals present, or suspected to be present, at the site based on
                                                                a review of site records. The next step is to define the physical/
                                                                chemical properties (e.g., vapor pressure and water solubility) and
                                                                route-specific toxicity for each listed chemical. The physical/chemi-
                                                                cal properties will be used in determining the migration potential
                                                                of each chemical.  The  migration potentials and route-specific
                                                                toxicity will be used to calculate the Chemical Factor Score.

                                                                Migration Potential Score
                                                                  Contaminants can be moved downward through the soil by water
                                                                as a result of infiltration induced by precipitation or irrigation
                                                                (leaching potential) or upward as a result of evaporation (evap-
                                                                oration potential). Finally, some contaminants are so tightly bound
                                                                to the soil that they are likely to remain near the surface of the
                                                                soil, sorbed to soil particles indefinitely (soil retention potential).
                                                                For the purposes of the HEPRM, Migration Potential Scores are
                                                                classified as  "high" (score = 3), "medium" (score = 2) and "low"
                                                                (score = 1) and are estimated based on the physical/chemical
                                                                properties of vapor pressure  and water solubility.
                                                                  The evaporation potential (EP) of a chemical is directly propor-
                                                                tional to its vapor pressure. The HEPRM determines three ranges
                                                                of vapor pressure which are consistent with volatility characteristics
                                                                associated with ranges of Henry's Law constants based on the two-
                                                                layer film or resistance concept of vapor transport from water.
                                                                An EP score is assigned based on these ranges of vapor pressure.
                                                                  The leaching potential (LP) of a chemical is directly proportional
                                                                to its water solubility. However, for the leaching rate of a chemi-
                                                                cal to be significant, the volatilization rate must not be greater than
                                                                the leaching rate. Therefore, the HEPRM assigns the LP score
                                                                based on the water solubility linked to a boundary vapor pressure.
                                                                The boundary vapor pressure was determined by comparing  the
                                                                vapor pressure of a variety of contaminants and the probability
                                                                of detecting those contaminants in groundwater and surface water,
                                                                as presented in the literature.
                                                                  The soil retention potential (SRP) of a chemical  is inversely
                                                                proportional to both water solubility and vapor pressure.  The
                                                                HEPRM defines the SRP as essentially the inverse of either  the
                                                                LP or the EP  (i.e., chemicals with either high LP or EP scores
                                                                will  have a low SRP score).

                                                                Toxicity Score
                                                                  The toxicity of  contaminants present at the site is classified
                                                                according to the Sax toxicity ratings2. The Sax toxicity rating
                                                                system classifies chemicals as being "severely toxic" (Toxicity Score
                                                                 = 3), "moderately toxic" (Toxicity Score = 2) or "slightly toxic"
                                                                (Toxicity Score = 1). For the purposes of the HEPRM, this rating
                                                                scheme is augmented by a fourth classification for carcinogenic
                                                                compounds (Toxicity Score  = 4).
                                                                                                   HEALTH ASSESSMENT     159

-------
  The toxicity of a chemical may vary depending on the route of
exposure (i.e., ingestion, inhalation, dermal contact). The route
of exposure is represented in the HEPRM by the individual human
exposure pathways (Table 1). Thus, it is necessary to identify route-
specific toxicity values for a contaminant whenever possible. The
Sax toxicity rating system provides route-specific toxicity scores
for some chemicals. The  DOH currently is  working towards
defining route-specific toxicity scores for additional chemicals based
upon a review of the lexicological literature.
   The third step in the methodology for selecting contaminants
of concern is to calculate Chemical Factor Scores as the product
of the  Toxicity Score  and the appropriate Migration Potential
Score. A total of 16 Chemical Factor Scores are required as input
to the HEPRM for each contaminant.

 Target Factor Score
   The Target Factor Score combines site-specific information on
 the size of exposed populations and the distance of exposed popu-
 lations from the site. In order to calculate the respective Target
 Factor Scores, it is necessary to estimate how  many people may
 be exposed via each exposure pathway, estimate their degree of
 exposure (i.e., resident versus non-resident) and characterize their
 distance and direction  from the site. The Target Factor Score for
 each target population is calculated as the product of a Popula-
 tion  Factor Score and a Distance Factor Score.

 Population Factor Score
   A  Population Factor Score ranging from 0 to 4 is assigned for
 each target group based on the total number of people potentially
 exposed within that target group. A target group is defined as the
 total number of persons potentially affected within a defined
 distance from the site (Fig. 3).  For example, a groundwater popu-
 lation factor is assigned by summing all individual target popula-
 tions potentially using groundwater as  a  water supply (e.g.,
 factories, schools, residential  neighborhoods)  within 1,000  m of
 a site to yield the total population within that  target group  (i.e.,
 target group within 1,000 m of a site). The groundwater popula-
 tion  factor for that target group is  assigned based on the  total
 population.

 Distance Factor Score
   A  Distance Factor Score ranging from 0 to 3 is assigned for each
 target group based on  how far the exposure point for that group
 is from the site. For example, the  groundwater distance factor
 would  be assigned based on the distance of the well from the site,
 not the distance of the user from the site. For example, a city well
 serving 10,000 people is located 1,000 m from the site but the closest
 resident served by that well lives 25,000 m from the site. The
 groundwater  distance  factor  would be assigned based  on the
 1,000 m  distance, not the 25,000 m distance.
   The Target Factor Score for each target group is calculated as
 the product of the population  factor and the distance factor. The
 sum  of the individual Target Factor Scores (i.e., for target groups
 at distances A, B, C,  etc.) is  divided by the maximum possible
 Target Factor Score for that pathway to yield  the final pathway-
 specific Target Factor Score for the site. A total of 17 pathway-
 specific target factor scores are calculated as input to the HEPRM.

 Probability Score
   The probability of contaminant transport by an exposure  path-
 way  may be determined by an observed incident of contaminant
 release or estimated from pertinent site factors. Figure 4 provides
 an overview of the procedure  for determining  probability scores.
   For  all exposure pathways, an observed incident of contamina-
 tion  or contaminant release results in a Probability Score of 1.0.
 Furthermore, if contamination is documented  at a known human
 exposure point (i.e., in a drinking water well) then the Probability
 Score is 1.1. This differentiation in the Probability Scores will  result
 in sites with documented contamination at known exposure points
 (i.e., in drinking water or in air in the home) ranking higher than
                       TFA
                          PFA
                                   DF
     where:   TF.  - Target Factor Score for  Target Croup A

                  Population Factor for Target Croup A

                  Distance Factor for Target Croup  A
          PFA

          DFA
   Total Target Factor Score   
-------
laminated drinking water well if both water sources have the same
level of contamination. Therefore, it is necessary to determine a
relative risk weighting factor for each pathway which reflects the
magnitude of human exposure and the resulting potential risk.
  The Weighting Factor for each pathway was determined by
estimating the daily intake, in ug/day, for each pathway based on
a series of common assumptions. Intake  by any pathway was
determined by multiplying a nominal concentration in the relevant
medium by an appropriate exposure coefficient3. Exposure coeffi-
cients were developed by assuming a reasonable worst-case fre-
quency of exposure (i.e., number of exposures/day), duration of
exposure (i.e., hours/exposure) and magnitude or rate of exposure
to the medium (e.g., average daily ingestion of water is 2 I/day)4.
Figure 5 provides an overview of the methodology for assigning
Weighting Factors. A Weighting Factor was assigned for each path-
way based on the following ranges of estimated daily intake:
          Intake (ug/day)

            > 1,000
            10 -  1,000
            0.1 - 10
            0.001 - 0.01
            > 0.001
                         Weighting Factor

                                1.0
                                0.8
                                0.6
                                0.4
                                0.2
                           CONFIRMED
                       INCIDENT OF CONTACT
                       AT OR RELEASE FROM
                            THE SITE
                                        ASSIGN VALUES TO
                                     APPROPRIATE ESTIMATION
                                           FACTORS
   CONTAMINATION
DOCUMENTED AT A KNOWN
   HUMAN EXPOSURE
       POINT
         I
                                       ESTIMATE PROBABILITY
                                      USING ASSIGNED VALUES
                                    AND APPROPRIATE EQUATION
                            Figure 4
  FREQUENCY
                  DURATION
                                   RATE
                  EXPOSURE
                 COEFFICIENT
                                   CONCENTRATION
                                     IN MEDIA OF
                                       CONCERN

1
r
ESTIMATED DAILY
INTAKE

                            WEIGHTING FACTOR
                                 ASSIGNED
                            Figure 5
                     Weighting Factor Scores
  A total of 40 Weighting Factors (i.e., one for each exposure path-
way) are required as input to the HEPRM. These Weighting Factors
are identified in the Methodology Report and will be the same for
all sites1.

OVERVIEW OF CALCULATION METHODOLOGY
  The score for each pathway is calculated as the product of the
appropriate Chemical Factor, Target Factor, Probability  and
Weighting Factor Scores.

Chemical Factor (CF) Score
  The Chemical Factor Score for each pathway is selected from
the contaminants of concern on the basis of exposure route. If no
route-specific toxicity score is available for a contaminant, then
the general toxicity score  should  be used.  Select the  highest
applicable Chemical Factor Score for each pathway from the con-
taminants of concern for the site. Divide the Chemical Factor Score
by 12 (the maximum possible score) and use the resulting ratio to
calculate the pathway score.

Target Factor (TF)  Score
  The appropriate Target Factor Score for each pathway is esti-
mated using site-specific information and the worksheets provided
in the Methodology Report1. Use the Target Factor Score as esti-
mated in the worksheets to calculate the pathway score.

Probability (P) Score
  The appropriate probability of contact or  transport score for
each pathway is estimated using site-specific information and the
worksheets provided in the Methodology Report1.  Use the proba-
bility as estimated in the worksheets for calculating the pathway
score.

Weighting Factor (WF) Score
  The Weighting Factor for each pathway is assigned on the basis
of the estimated worst-case daily intake which could result from
exposure via each pathway.  Use the weighting factor assigned for
each pathway in the Methodology Report1 to calculate the path-
way score.
  The maximum score for pathways  varies from  0.2  to 1.1 and
is calculated as follows:
                                                                     Pathway Score = CF/12  x TF X P WF
                                                                                                                   (2)
  The respective media scores (i.e., air, soil, groundwater and
surface water) are calculated by summing the individual pathway
scores within each medium. The overall site score is calculated by
summing the respective medium scores. The overall site score is
then divided by 21 (the maximum possible score for any site) and
multiplied by 100 to yield a final score on a scale of 1 to 100.

CONCLUSIONS
  The HEPRM currently is being used by DOH to rank approxi-
mately 750 sites. The results of the ranking will be used by DEC
and DOH in the allocation of state funds for hazardous waste site
cleanups.
  The HEPRM fully considers the three human exposure routes
(i.e.,  inhalation, ingestion and dermal contact) and creates the
mechanisms to assess all of the significant environmental factors
leading to human exposure. Thus, the  HEPRM is designed to be
a useful tool for planning future investigations and/or remedial
activities at abandoned hazardous waste sites.

REFERENCES
1. Smith L.A. and C.D. Patrick. "Human Exposure Potential  Ranking
   Model for Hazardous Waste Sites: A Methodology Report." New York
   State Department of Health. Albany, NY, TR-847-2C. Contract No.
   HRI-814-2165A, with  ICAIR Life Systems, Inc., 1986.
2. Sax, N.I.  "Dangerous Properties of Industrial Materials." Van Nostrand
   Rheinhold Co., New York, NY, 5th Edition, 1979.
3. Schultz, H.L., W.A. Palmer, G.H. Dickson, et al. "Superfund Expo-
   sure Assessment Manual." Draft. U.S. EPA, Office of Solid Waste and
   Emergency Response,  Washington, DC. OSWER Directive 9285.5-1.
4. Hawley, J.K. "Assessment of Health Risk from Exposure to Contami-
   nated Soil." Risk Analysis S, 1985, 289-302.
                                                                                               HEALTH ASSESSMENT     161

-------
            Hazardous Waste Site Health  and  Safety  After  OSHA
                                                 Martin S. Mathamel
                                      COM Federal  Programs Corporation
                                                  Fairfax,  Virginia
ABSTRACT
  This paper focuses on the new OSHA rules for hazardous waste
site and emergency response operations and what federal, state and
local government organizations as well as consultants, contractors
and industry need to do to comply with them.

INTRODUCTION
  With increasing government regulation over the past several
years,  hazardous waste site (HWS) worker health and safety
programs have dramatically increased in complexity. In the early
1980s,  the U.S. EPA began setting HWS safety standards. These
standards were published in "Health and Safety Requirements for
Workers Engaged in Field Operations" (1). This document was
the basis for the  first widely distributed set of HWS safety
standards, the "Standard Operating Safety Guidelines"2. In  1985,
NIOSH/OSHA and the USCG began to become involved in HWS
safety, resulting in the release of "Occupational Safety and Health
Guidance Manual for Hazardous Waste Site Activities"3—the so-
called "four agency manual." This manual did not have the  force
of law  behind it and hence was considered to be only a guideline.
  The  passage of SARA in 1986 brought congressional mandate
".. .to protect hazardous waste site workers whenever they deal
with hazardous wastes." SARA tasked OSHA with promulgating
regulations  to enforce SARA Section 126—"Worker Protection
Standards." In December 1986, OSHA released the "Hazardous
Waste  Operations  and  Emergency  Response  Interim  Final
Rules"4, which took effect in March 1987 and remain in effect
until October 1988.  Combined with the final rules to be released
in October  1987 (which take effect October 1988), OSHA has
promulgated a set of enforceable and extremely comprehensive
HWS health and safety regulations that have the force of law
behind them.  These rules require:

  Site  characterization
  Site  analysis training
  Medical surveillance
  Work practices
  Engineering controls
  Monitoring
  Informational programs
  Personnel protective equipment
  New technology programs
  Decontamination
  Emergency response

  These requirements  are generally met as part of a comprehen-
sive health, safety  and training program  for HWS workers.
Implementation is usually via site specific health and safety  plans
and a comprehensive worker certification program. Both are
discussed below.
  Note  that although 29  CFR 1910.120 applies  directly to
hazardous waste site operations, the applicable portions of 29 CFR
1910 (General Industry) and 29 CFR 1926 (Construction) standards
also apply.
  In the following text, references in parentheses refer to specific
sections  of the interim final standard.

APPLICABILITY
The OSHA standard applies to:
• Workers involved in CERCLA/SARA hazardous waste site oper-
  ations, such as remedial investigation/feasibility study (RI/FS)
  and cleanup  activities
• Workers involved in RCRA major corrective actions (RI/FS and
  cleanup activities); RCRA major corrective actions are consi-
  dered  to be CERCLA/SARA  activities for the purpose of the
  standard
• Activities conducted at 40 CFR 264 and 40 CFR 265 RCRA treat-
  ment,  storage and disposal facilities (only Paragraph (0) of the
  standard applies). These activities are beyond the scope of this
  paper  and are not discussed. The reader is referred directly to
  29 CFR 1910.120.
• Hazardous waste site operations designated by  state or local
  government.
• Emergency response operations for hazardous substance releases.
  Emergency response operations are beyond  the scope of this
  paper  and are not discussed; the reader is referred directly to
  29 CFR 1910.120.

  Note that operations at municipal or sanitary landfills that do
not handle hazardous waste are not covered by the  standard.
  The  standard specifically  covers the  following  worker
populations:

• General industry, federal government and local government
  workers including the following 24 states/territories: Alaska,
  Arizona, Connecticut (for state and local government workers
  only), Hawaii, Indiana. Iowa,  Kentucky, Maryland, Michigan,
  Minnesota, Nevada, New Mexico,  New York (for state and
  local government workers only), North Carolina, Oregon, Puerto
  Rico,  South Carolina, Tennessee, Utah, Vermont, Virginia,
  Virgin Islands, Washington and Wyoming.
• The remaining 28 states without Federal OSHA-approved state
  plans  will be required to follow the final rule promulgated by
  the U.S. EPA.

  For all practical purposes, due to liability considerations, the
162    HEALTH ASSESSMENT

-------
standard applies to all hazardous waste site activities performed
by general industry including professional engineering and scien-
tific firms.
  Every portion of the interim final standard is in effect  as of
Mar. 16, 1987 [ref. para. (p)].  The final standard is expected to
take effect in October 1988.

CERCLA/SARA FIELD OPERATIONS
  The major thrust  of the standard is CERCLA/SARA field oper-
ations, such as conducting an RI/FS or cleanup. As stated above,
RCRA  major corrective actions fall under this  portion of the
standard.  Organizations have two broad areas of responsibility
under this part of the standard. One is the generation of an
appropriate site-specific health and safety plan (HSP) that ensures
compliance with  the  set of standard site practices  required by
OSHA. The second is compliance with worker field  certification
standards.

HEALTH AND SAFETY PLANS
  The HSP is the major vehicle which establishes site safety pro-
cedures. By effectively implementing the HSP, the procedures can
be enforced. This section discusses the minimum information that
needs to be contained in the HSP and specific procedures that need
to be followed.

On-Site Personnel
  HSPs must list on-site personnel by task who are field certified
and meet the medical, training  and respiratory fit testing require-
ments discussed in  the Worker Field Certification section below.

Preliminary Off-site Evaluation
  HSPs must fully characterize the site  hazards  [ref. para. (c)].
The preliminary off-site evaluation must include determinations
of the following:
• Site location and size
• Site topography
• Site accessibility  by air and  roads
• Hazardous substances involved or expected at the site and their
  chemical and physical properties
PRELIMINARY Of FSITE EVALUATION
[REVIEW EXSTNC BACKGROUND
^FORMATION)


           SUFFICIENT
       INFORMATION FOR A
     POSITIVE IDENTIFICATION
     AND QUANTIFICATION OF
        HAZARDS OR
        SUSPECTED
         HAZARDS?
     SELECT APPROPRIATE PPE
   FOR STTE ENTRY AND SUBSEQUENT
        WORK ACTIVITIES
                                     POSITIVE IDENTIFICATION
      SITE ENTRY (LEVEL D-A)
    (AIR MONITOR WITH Dfll OR SPl)
      COMFIFM PROPER PPE
   FOR SUBSEQUENT WORK ACTWmES
   IMITATE WORK ACTIVITIES AND
   ONGOH5 MOMTOFaNO PROGRAM
         (DRIVSPl)
                            Figure 1
                Site Characterization and Analysis
• Pathways for hazardous substance dispersion
• Description of response activities tasks/worker job functions
• Planned duration of worker activity
• Present status and capabilities of emergency response teams for
  worker on-site emergencies

Initial Site Entry
  An initial site entry is performed in order to verify the prelimi-
nary off-site evaluation.
  Based on the preliminary off-site evaluation, personnel protec-
tive equipment (PPE) is selected and specified in the HSP for initial
site entry. The flow chart presented as Figure 1 can be used to select
PPE. For initial site entry:

• If the preliminary off-site evaluation does not provide enough
  information, then  Level B PPE is used for initial site entry
• Escape packs are worn or located at the work station for Level
  C or  Level D initial site  entries

  Note that initial site entry is defined as the first visit at a  site.
Entry into sites that have had continuing, ongoing activity or are
fully characterized are not considered to be initial site entry.  The
preliminary off-site evaluation takes these factors into account.

Site-Specific HSPs
  Site-specific HSPs [ref. para, (i)] must specify:

• Name of key personnel and health and safety
• Personnel task/operation safety and health risk analysis
• Worker training
• Personal protective equipment to be used
• Medical surveillance requirements
• Frequency and types of air monitoring, personnel monitoring
  and sampling techniques
• Site control measures
• Decontamination procedures
• Site standard operating procedures (SOPs)
• Contingency plans
• Confined space entry procedures

Site Health and Safety  Coordinator
  Each site normally has a designated site health and safety coor-
dinator  (SHSC) who is in charge of site safety. The SHSC enforces
the provisions of the HSP and stops work if site conditions become
unsafe.  In addition, the SHSC must hold frequent safety briefings
and must regularly perform job-site safety inspections.

Site Control Program
  HSPs must include a site control program [ref. para, (d)]  that
includes the following:

• Site map
• Specified site work zones, e.g., exclusion, decontamination and
  support zone
• Use of a buddy  system
• Site communications
• Standard operating procedures (SOPs) for safety practices

Exposure Monitoring
  HSPs must specify monitoring and sampling procedures  [ref.
para, (h)] for measuring exposures on sites.

• Air monitoring is used to identify and quantify airborne levels
  of hazardous substances in order to determine and/or confirm
  the appropriate level  of PPE.
• Air monitoring also is performed for IDLH conditions and other
  hazards, such as flammable/oxygen deficient atmospheres, toxic
  levels, radioactive  materials, etc.
• Direct reading instruments (DRIs) generally are used to perform
  the air  monitoring tasks  referenced above.
• Air sampling utilizing sampling pumps (SPs) and collection media
  is performed for longer-term site operations involving hazards
                                                                                                HEALTH ASSESSMENT     163

-------
  like asbestos when site personnel remain  in a location  long
  enough to get usable analytical results from a laboratory.

  In addition, monitoring and sampling are  performed:
  - when work begins on a different portion of the site
  - when handling unidentified  contaminants then identified
  - when a different type of operation is initiated
  - when handling leaking drums or working in obvious liquid
    contamination

• Monitoring performed during initial site entry is termed initial
  monitoring.  Monitoring performed after  initial  site entry is
  termed continuing or periodic monitoring.

Emergency  Planning
  HSPs must provide for emergency response planning [ref. para.
(1)1.
• An Emergency Response Plan (ERP) is developed and im-
  plemented to handle anticipated on-site emergencies prior to the
  commencement of hazardous waste operations. The ERP covers
  both on-site and off-site emergencies. The ERP addresses:
  -pre-emergency planning
  -personnel roles, authority, training and communication
  -safe distances and place of refuge
  -site security and control
  -evacuation  routes and procedures
  -decontamination
  -emergency medical treatment and first aid
  -emergency alerting and response procedures
  -critique of response and followup
  -PPE and emergency equipment

• In addition, the following elements are included:
  -site topography, layout and prevailing weather conditions
   -state, local  and  federal reporting procedures for specific
   incidents

  The ERP is a separate section of the HSP and is rehearsed
regularly as part of the overall training program.

Exposure Control
  Exposure to hazardous materials is held below safe levels by a
combination of engineering controls, work practices and personal
protective equipment (PPE) [ref. para. (g)]. The HSP must  state
how exposure control  is  to be accomplished.

• Engineering controls, work practices and PPE are used to reduce
  and maintain  worker  exposure to or below  the  Permissible
  Exposure Limit (PEL).
  -organizations may not implement a schedule of worker rotation
  as a means of compliance with the PEL
 -for the purpose of the standard, the PEL is defined as the Per-
  missible Exposure Limit (OSHA) or the Threshold Limit Value
  (TLV) (ACGIH) or other accepted  or  established  exposure
  criteria, whichever is lower

• PPE is based on hazards and/or potential hazards.
  -the PPE selected must conform to the U.S. EPA levels "A",
    "B", "C" or "D"

• The PPE program addresses the following:
  -site hazards
  -PPE selection
  -work mission duration
  -PPE maintenance and storage
  -PPE decontamination
  -PPE training and proper fit
  -PPE donning and doffing procedures
  -PPE inspection
  -PPE in-use monitoring

164    HEALTH ASSESSMENT
  -evaluation of PPE program effectiveness
  -limitations during temperature extremes

Materials Handling
  Materials handling practices must not increase site hazards. The
HSP must specify material handling procedures. In particular,
handling of drums and containers that might contain hazardous
materials must meet the following criteria:

• Drums/containers  must  be  inspected for integrity  prior to
  moving.
  -unlabelled drums/containers shall be considered hazardous sub-
   stances and handled accordingly
  -site operations  must  minimize  amount of drum/container
   movement
  -prior to moving drums/containers,  all workers must be noti-
   fied of hazardous substances.
  -where major spills may occur, a spill containment program must
   be implemented
  -a ground-penetrating  system must  be used to locate buried
   drums or  other  containers
  -fire extinguishing equipment must be present  to handle small
   fires

• The following procedures must be followed during drum/con-
  tainer operations:
  -airline respirator systems must be protected from contamina-
   tion  or physical damage
  -personnel not directly involved must keep a safe distance
  -if workers must work near  or adjacent to drums/containers
   being opened, a suitable shield  must be placed between the
   worker and the drums/containers
  -operational controls and safety equipment must be located
   behind a explosion-resistant barrier
  -all tools must be spark proof
  -workers must not stand on or work from drums/containers

• The following special precautions must be taken when handling
  drums/containers which  are suspected of containing shock-
  sensitive materials:
  -non-essential personnel must be evacuated
  -explosive containment devices/shields must be provided
  -a worker alarm system must be used
  -continuous communications must be maintained
  -pressurized drums/containers must be routinely moved
  -routine  laboratory packs must  be considered  shock sensi-
   tive/explosive

Decontamination Procedures
  Decontamination procedures must be established [ref. para. (k)l
for personnel and equipment. The HSP specifies decontamination
procedures for personnel and equipment. In addition, decontami-
nation procedures must be monitored by the SHSC to determine
their effectiveness. Commercial laundries which clean protective
clothing or equipment must be informed of the potentially harmful
effects.

Illumination
  When work is done at night, sufficient illumination (enough to
read a newspaper with difficulty) must be provided [ref. para. (m)).

Sanitary Facilities
  Sanitary facilities must  be provided at each site [ref. para. (n)).
These facilities must include drinking water, wash-up water, fire-
fighting water and toilet facilities. For  short-term operations, an
automobile in which the workers can drive to a toilet facility meets
the standard. The HSP needs to discuss how sanitation is to be
provided.

Excavation Procedures
  Excavation practices followed on-site must meet  the OSHA  Con-

-------
struction  Standards (29 CFR 1926) which include shoring and
sloping trenches in soft soil to prevent accidental burials of crew
members. These practices must be detailed in the HSP.

WORKER FIELD CERTIFICATION PROGRAM
  As previously discussed, compliance with the HSP is the major
mechanism by which safe practices are established and enforced
on the HWS. However, before workers  can participate in HWS
activities, they must be field certified.
  Workers assigned to hazardous waste sites must be medically
monitored,  receive adequate training and  be fit tested for the
specific respirator assigned to  them  before  they can be  field
certified.

Medical Surveillance
  A medical surveillance program [ref. para, (f)] must be provided
for all workers who may be exposed to hazardous substances at
or above the permissible exposure limit (PEL) without regard to
the use of respirators for at least 30 days/year and/or workers who
use respirators for at least 30 days. Note the following:

• Frequency of medical examination
  -prior to assignment
  -at least once every 12 months
  -upon termination
  -as soon  as possible  after a  worker has developed signs/
   symptoms from possible exposure

• Content of examination
  -a medical and work history with special emphasis directed to
   symptoms related to handling hazardous  substance
  -fitness for duty including  the  ability to wear required PPE
  -the  specific content to be  determined by  licensed  examining
   physician

Training Program
  Each worker who works on a hazardous  waste site must com-
plete a 40-hr basic health and safety training program [ref. para.
(e)] that includes, as a minimum;

  On-site health,  safety and hazards present
  Use of  PPE
  Safe work practices
  Safe use of engineering controls and equipment
  Medical surveillance requirements
  Site health and  safety plan  development

  The basic  training course generally meets  the informational
program requirement of the standard. In addition, the training
requirement includes a minimum of 3 additional days  of on-the-
job training (OJT) under supervision.
  SHSCs  receive the 40 hr of basic training,  plus the 3 days of
OJT and  8 hr additional training on managing site health and
safety.
  Lastly,  all workers  must complete 8 hr  of refresher training
annually.

Respirator Program
  Workers are required to be trained in  proper respirator usage
and to be  respirator fit tested. This includes compliance with the
respirator  standards contained in the 29 CFR 1910 General Industry
Standard.

JOB-SITE INSPECTIONS
  The only means of effectively enforcing the  HSP and field cer-
tification  program is to perform a job-site inspection.
  OSHA  requires frequent job-site inspections to ensure com-
pliance  with health and safety regulations.  The SHSC normally
is required to perform the job-site inspection.  In addition, an
individual not normally associated with the site should periodically
perform an unannounced audit. The form that appears as Figure
2 can be used as a check list for a site safety audit.

 SITE HEALTH AND SAFETY AUDIT FORM
 Site Name / Location / Number	
 Site Manager	

 Site Health and Safety Coordinator (SHSC).
 Audit Performed By_
                                        .Date.
                                 REQUIRED

                                   ()


                                   ()
 Signed HSP Onsite/Available
 HSAM Onsite/Available
 Health and Safety Signoff Form Completed
 For All Onsite Personnel
 MSDSs Onsite/Available                 ()
 Designated SHSC Present               ()
 Siie Safety Briefing Conducted            ()
 Onsite Personnel Meet HSAM Requirements  ()
 For Medical. Fit Test, and Training
 (including Subcontractors)
 Compliance With Specified LOPs
                                             IN COMPLIANCE?
                                             YES      NO
                                             ()
                                             ()
                                             ()

                                             ()


                                             ()
                                             ()

                                             ()
                                             ()

                                             ()

                                             0
                                             ()

                                             ()
                                             ()
                                             ()

                                             ()


                                             ()
                                   ()
 Equipment Specified In HSP Available       ()
 and In Working Order
 Adequate Equipment Inventory Available     ()
 Monitoring Equipment Calibrated          ()
 and Calibration Records Available
 Exclusion Zone, Decontamination Zone,     ()
 and Support Zone. Enforced
 Proper Decontamination Procedures        ()
 Emergency Phone Numbers Posted        ()
      24-Hour Number Posted            ()
 Emergency Route To Hospital Posted       ()
 Local Officials Notified                  ()
 Responsible Personnel Know How         ()
 To Operate Monitoring Equipment-
 Equipment Manuals Available
 Environmental/Personnel Monitoring        ()
 Performed As Specified In HSP
 First Aid Available                     ()        ()        ()
 Special Procedures Implemented          ()        ()        ()
          Record Comments ana Detail Non-Compliances On  Back 01 Form
                           Figure 2
               Site Health and Safety Audit Form
CONCLUSION
  OSHA has promulgated an extremely comprehensive set of regu-
lations governing HWS operations that have the force of law behind
them. As with  any OSHA  standard, it represents  the minimum
practices with which organizations must comply. Organizations
involved in HWS operations must have the foresight to have state-
of-the-art health,  safety and training programs that protect their
HWS workers and  comply with the standard or, indeed, exceed
the standard.  This is particularly true  today as  it  becomes
increasingly difficult for organizations working on HWSs to obtain
insurance to protect their assets in a legal environment that is
experiencing a rapid increase of personal injury suits. Statistics that
indicate the accident/injury rate for HWS workers to be 2.5 times
greater than for manufacturing industries support this conclusion.

REFERENCES

1.  U.S. EPA. "Health and Safety Requirements for Workers Engaged in
   Field Operations," U.S. EPA, Washington, DC, 1981
2.  U.S. EPA. "Standard Operating  Safety Guidelines:" U S  EPA
   Washington, DC,  1984
3.  NIOSH, OSHA, U.S. EPA and USCG. "Occupational Safety and
   Health Guidance Manual for Hazardous Waste Site Activities " NIOSH
   OSHA, U.S. EPA and USCG, Washington, DC, 1985
4.  OSHA, "Hazardous  Waste Operations and Emergency Response
   Interim Final Rules" 29 CFR 1910.210
                                                                                                 HEALTH ASSESSMENT     165

-------
                       Assessing  Risk  from Dermal  Exposure  at
                                        Hazardous Waste  Sites

                                                 Elizabeth A. Ryan
                                               Elizabeth  T. Hawkins
                                                 Brian  Magee, PhD
                                                   Susan  L. Santos
                                                   E.G.  Jordan Co.
                                                   Wakefield, MA
ABSTRACT
  The current methodology for estimating body dose levels from
dermal contact with contaminated soils is to use an absorption
factor of unity when empirically derived absorption factors are
unavailable1. Using a value of  100% is an overly  conservative
approach which leads to the gross overestimation of the dermal
absorption of most compounds. This paper provides one approach
for refining dermal absorption factors based on the  current toxi-
cology literature and chemical and physical considerations of the
contaminants. Absorption factors are developed for the three major
chemical classes; volatile organic compounds (VOCs); semi-volatile
organic compounds (SVOCs) including pesticides; and inorganic
compounds.

INTRODUCTION
  Estimating the potential human health risks posed by  con-
taminants at hazardous waste sites requires an  assessment of all
the possible routes of contaminant exposure. This process can be
very complex, as a large number of chemicals  often are present
and there often are multiple exposure pathways involving oral,
dermal and inhalation  routes of entry. Traditionally, when per-
forming a public health risk assessment, the major emphasis has
been on evaluating the ingestion of contaminated groundwater,
as this route of exposure  tends to generate  the greatest public
concern and is often the easiest  to evaluate. However, there are
many sites where the potential public health risks from direct con-
tact with contaminated soils or sediments need to be evaluated.
These sites include areas with unrestricted or limited  restriction to
access and/or sites where sensitive populations, such as children,
may come into contact with contaminated soils.
  Soil and sediment contamination is a common problem at most
hazardous waste sites  resulting from the indiscriminate disposal
practices or direct  dumping of contaminants onto surface soils.
Subsequent contaminant migration and contaminant transforma-
tion and fate processes often result in contamination to both the
surface  and subsurface soils  and sediments.  Exposure to  this
medium may result from one or more of the  following routes of
exposure: soil ingestion due either to inadvertent ingestion or  pica,
inhalation of  fugitive dusts and direct contact through dermal
exposure to soils.
  Although each of these three routes of exposure is important,
the focus of this paper is on assessing risk from dermal exposure
only. This route of exposure poses some unique challenges to risk
assessors as the toxicology literature rarely provides information
on the health effects caused by dermal exposure. In addition, bio-
availability considerations  often complicate the estimation of
incurred human doses posed by contact with contaminants in soil
or sediments. Because of these limitations, some risk assessors are
reluctant to quantitatively evaluate exposure from direct contact.
We, however, feel that it is possible to  refine certain exposure
parameters using the available  toxicity information  to provide
sound risk estimates that can be used to  make risk management
decisions.
  The subsequent sections of this paper  provide an overview of
the information needed to assess risk from dermal contact with
soils and a discussion of how the available  information can be used
to derive absorption factors for contaminants in a soil matrix.

NECESSARY AND AVAILABLE INFORMATION
  The actual risk from dermal contact with contaminated soils will
be a  function of the contaminant concentration in the soil, the
amount of soil contacted, the frequency of contact and the amount
of contaminant absorbed through the skin. To provide a quan-
titative estimate of the risk from this route  of exposure, numeri-
cal values must be assigned to each of these exposure parameters.
When combined, these parameters yield the "body dose level  of a
contaminant expressed in mg/kg-das. The body dose level can
either be averaged over a lifetime and multiplied by the  cancer
potency factor to provide an estimate of the incremental  cancer
risk or divided by the most appropriate standard, criterion or guide-
line to provide a non-carcinogenic risk ratio.
  While  values for contaminant concentrations can be easily
obtained from analytical data, the other exposure parameters are
more difficult to quantify as these parameters often are based on
site specific factors such as the assumed activity and  behavior
patterns of the exposed population and/or the physical site condi-
tions. Researchers such as Renate Kimbrough, John Hawley and
John Schaum have developed methodologies for quantifying many
of these exposure parameters including the amount and frequency
of soil contacted per exposure. However, the  literature contains
little human or animal toxicology data on the  dermal absorption
of contaminants from a soil matrix. Dermal absorption informa-
tion is available for only a few of the many contaminants expected
to be present at a hazardous waste site, and much of the available
information is based on a contaminant's absorptive behavior in
a liquid or vapor phase.
  The limited quantitative data available indicate that  dermal
absorption of contaminants in a liquid phase occurs at rates less
than 100%. The data also show large variations in the estimated
absorption values of most contaminants tested due to the many
factors which influence the absorption of contaminants across the
skin including location of application, skin condition, concentra-
 166    HEALTH ASSESSMENT

-------
tion and bio-availability2'3. While this information is useful for
providing qualitative information  regarding  a  contaminants
absorption potential, it does not provide the information needed
for assessing dermal absorption of contaminants in soils. Values
based on the absorption of a contaminant in a liquid or vapor phase
will overestimate the absorption of the same contaminant in a soil
matrix as the absorption value does not account for the adsorp-
tion/desorption processes which govern the contaminants bio-
availability.  In addition, the contaminant may be in a different
chemical form in a soil  matrix than in a liquid or vapor phase,
thereby altering its propensity to be absorbed.
  Since data generally are not available for studies of contaminants
adsorbed to  soils, scientific judgments  must be used to assess the
potential for such contaminants to enter the body and cause adverse
health effects. The following discussion details an approach for
assessing dermal absorption and deriving appropriate absorption
factors.

METHODOLOGY
  McLaughlin2 discusses two approaches which can be used to
estimate the dermal absorption of different contaminants. The first
approach uses a total absorption factor (e.g., 70% or 80%) to
modify the exposure concentration. These factors are estimated
based on the toxicology literature and the chemical and physical
characteristics of the contaminant. This approach does not account
for contact time or the particular kinetics that may apply to con-
taminant transport across the skin.
  The second approach uses an absorption rate which is coupled
with a contact time to provide an estimate of the amount of con-
taminant  penetrating the skin.  The  absorption  rate is either
measured directly or estimated based on the  chemical properties
of  the contaminant  (i.e., Kow,  Koc  and diffusion constants).
Because of the limitations and uncertainties inherent in determining
the contact  time, kinetics and/or other variables required for
deriving an absorption rate, this methodology is considered to be
limited in its usefulness for assessing dermal absorption. As such,
the focus of this paper is on assessing dermal absorption using total
percentages  or absorption factors.
  The most conservative approach for determining body dose levels
in the absence of empirically derived absorption factors is to assume
complete absorption of a contaminant across the skin barrier. The
rationale for this approach is that it will provide the most conser-
vative estimate of the incurred body dose level and thus produce
the most conservative risk estimate. However, our experience in-
dicates that  assuming 100% absorption leads to a gross overesti-
mation of the actual dermal absorption of most compounds and
thus does not provide a realistic estimate of risk even under worst-
case exposure conditions. A more reasonable approach for deter-
mining body dose levels is to develop an absorption factor based
on  the contaminants' physical and chemical properties. This esti-
mation can be accomplished by extrapolating from applicable scien-
tific and pharmacokinetic data. Using this approach,  it has been
possible to assign more realistic estimates of the dermal absorp-
tion of a contaminant and thus decrease the uncertainty of the es-
timated body dose level used to assess risk.
  The methodology proposed here does not assign chemical specific
absorption factors but rather provides a range of values for the
three major chemical  classes. These  ranges are  considered to
encompass the actual dermal absorption of a specific contaminant
occurring under various exposure conditions. Using  a range of
values provides a limited sensitivity analysis by which uncertain-
ties can later be evaluated.
  Absorption factors are developed for the three major chemical
classes; volatile organic compounds (VOCs), semi-volatile organic
compounds  (SVOCs) and inorganic compounds. Grouping and
assigning  absorption ranges by chemical class implies that these
compounds  will  behave  similarly  both environmentally  (bio-
availability) and in then" propensity to diffuse through the skin
barrier. While this is an oversimplification, the limited amount of
chemical specific absorption information precludes further refine-
ment even if chemicals were classified in a more specific manner
(i.e., by Kow, Koc, molecular weight, etc.). The absorption factors
proposed here are used by Jordan when chemical specific data are
not available. These factors are expected to become more refined
as more pharmacokinetic data become  available.
  The proposed absorption rates for each chemical class are based
on reviews of the current literature and refined based on the phys-
ical and chemical properties of the contaminant. For volatile
organic compounds, a range of 10-25 % dermal absorption is con-
sidered to be representative of the absorption potential of these
compounds bound in a soil matrix. This range is based on pharma-
cokinetic data which show that some volatile compounds have high
percutaneous absorption when applied on the skin in solution2'3.
Analytical data for selected VOCs show  the dermal absorption to
be as high as 50% of the total exposed concentration3. These
studies, however, are based on the absorption of a contaminant
exposed  to  the skin in  a pure  or  dilute solution. The  dermal
absorption of these chemicals from a soil matrix is expected to be
lower than  that observed  in  a  liquid phase as a result of the
decreased bio-availability  of the contaminant.  However,  the
decreased bio-availability is not considered that great as VOCs do
not demonstrate a strong adsorption to soil particulates. This fact,
combined with the relatively high lipid solubility, suggests that these
compounds  may  be bio-available in a  soil  matrix. Absorption
factors between 10 and 25% are therefore considered to be applica-
ble for this  class of compounds.
  Some limited chemical specific information is available for the
absorption of some semi-volatile compounds. For example, the
dermal absorption of 2,3,7,8-TCDD in soil has been observed in
animal models to be between 0.07 and 3 % of the total soil-bound
TCDD4. Kimbrough et al.,5 concluded that between 1 and 10 %
of TCDD present in soil that comes in contact with human skin
may be absorbed. An absorption factor of 5  % has been recom-
mended by the U.S. EPA for assessing dermal exposure to PCBs
in soils6.
  Studies based on the topical administration of pesticides show
dermal absorption percentages ranging from 0.3 (Diquat) to 74 %
(Carbaryl) with most compounds being between 5 and  15%2 The
dermal absorption of these pesticides contained in a soil matrix
are expected to be lower. The physical and chemical properties of
SVOCs and pesticides indicate that these compounds  are tightly
bound to soils and thus are not readily bio-available. Although
many of these compounds are lipid soluble, the low bio-availability
of these compounds in  a soil matrix suggests that the  dermal
absorption will be low. An absorption range  of 1 to 10% is con-
sidered representative of the absorption of these compounds.
  Inorganic contaminants are considered to have the lowest dermal
absorption potential of  the three chemical  classes. These com-
pounds bind strongly to soils, greatly reducing their bio-availability.
In addition, their ionic speciation and/or formation of hydroxide
or large organic complexes further inhibits their absorption poten-
tial. A range of  absorption factors considered appropriate for
assessing exposure to inorganic compounds is 0.1 to  1%.
  The literature reports absorption ranges for various metals in
solution up  to several percent. Metal behavior in solution differs
greatly from its behavior in a soil matrix as the transformation
and fate processes are much more complex in a soil system. Most
metals form strong, stable bonds with other soil constituents which
effectively reduces the concentration of the dissolved metal. For
example, arsenic is known to form tight bonds with three commonly
occurring metals in soils;  iron, manganese and aluminum. The com-
plex, iron-arsenate, is poorly soluble in water and at equilibrium
the concentration of soluble arsenic is  only  1  x  10~;; mole/1.
The presence of iron or other complexing agents such as humic
and fulvic acids, will greatly decrease the bio-availability of most
inorganic compounds as long as the environmental conditions favor
the complex formation. These factors support the relatively low
range  of absorption factors  presented above.
                                                                                              HEALTH ASSESSMENT     167

-------
   We have found that in  addition to the toxicology literature
 analytical data can be used to refine absorption factors for inor-
 ganic compounds. The results from the U.S. EPA's EP-Toxicity
 test can be used to provide chemical and site-specific information
 regarding  the  bio-availability of certain  inorganic compounds
 (arsenic, barium, beryllium, cadmium, chromium, copper, lead,
 mercury, nickel, silver, selenium and zinc). This test provides a
 measure of the amount of metal present  in an acid extractable
 fraction7.  Because the test conditions are much  more rigorous
 than actual exposure conditions, the  EP-Toxicity results provide
 an upper bound of the amount of contaminant which is potentially
 bio-available for dermal absorption.
   The test results can be used to modify the exposure concentra-
 tion used in calculating incurred body dose levels by providing the
 concentration  of the contaminant in solution and/or the bio-
 available fraction. The dermal absorption factor applied to this
 exposure scenario does not need to account for the bio-availability
 of the contaminant  as is necessary  to assess exposure to con-
 taminants bound to  soils. Instead, the absorption factors can be
 based directly on analytical results. The scientific literature shows
 that between 1 and 5 ฐ7o of inorganic compounds can be dermally
 absorbed when exposure occurs in the liquid phase. Thus, this range
 of values can be used  to estimate the dermal absorption of inorganic
 compounds in this bio'available form.

 CONCLUSION
   Ranges of absorption factors suitable for assessing the incurred
 body dose levels of contaminants resulting from dermal contact
 with contaminated soils are presented for the three major chemi-
 cal classes. These values are derived based on the available toxi-
cology data and reflect the contaminants' physical and chemical
properties. For volatile organic compounds, the absorption factors
range  from 10  to 25%;  for semi-volatile organic compounds
including pesticides, the absorption factors range from 1 to IQVt;
and for inorganic compounds, the absorption factors range from
0.1 to  1%.
   These ranges provide a means for more accurately estimating
direct contact  exposure with soils. As more information on con-
taminant absorption becomes available, these absorption factors
could be further refined. Eventually it is hoped that more research
will be performed in this area to provide chemical specific absorp-
tion  factors for the contaminants  most commonly  found at
hazardous waste sites.

REFERENCES
I. U.S.  EPA Superfund Public Health Assessment Manual. Office of
   Emergency and Remedial Response, Washington, DC, Dec.  1986. EPA
   540/1-86/060.
2. U.S.  EPA "Review of Dermal  Absorption." Office of Health and
   Environmental  Assessment,  Washington,  DC,  Oct.  1984.
   EPA/600/8-84/033.
3. OCA Corporation "Assessment of Dermal Absorption of Contaminants
   in Drinking Water." Prepared for the U.S. EPA Office of Drinking
   Water, April 1984. Contract No. 68-02-3168.
4. Poiger, H. and Schlatler, C. "Influence of Solvents and Adsorbanu
   on Dermal and Intestinal Absorption of TCDD". Food and Cosemtic
   Toxic. 18. 1980, 477-481.
5. Kimbrough. R.D., FaJfc. H.. Stehr, P. and Fries, G. "Health Implica-
   tions of 2,3,7,8-Tetrachlorodibenzodioxin (TCDD) Contamination of
   Residential Soil." J. Toxicol Environ. Health 14,  1984. 47-93.
168    HEALTH ASSESSMENT

-------
                      On-Site Water  Treatment  Under  Superfund
                                      Economical and  Effective
U.S.
                                                   Jeffrey S. Clark
                                             MAECORP Incorporated
                                              Grand Rapids,  Michigan
                                                 Peter D.  Neithercut
                                           Environmental  Protection Agency
                                               Eastern Response Unit
                                                Grosse He,  Michigan
ABSTRACT
  On Dec. 3, 1986, an emergency removal action was initiated by
the U.S. EPA at the Commercial Oil Services site in Oregon. Ohio.
MAECORP Incorporated was issued a delivery order to design,
fabricate, install and operate a mobile wastewater treatment system
consisting of oil/water separation, chemical precipitation, clarifi-
cation, sand  filtration and carbon adsorption.
  The decision was made to treat on8 site as opposed to transporting
to an off-site disposal facility for several reasons: better control
over the treatment, less risk due to handling and transporting and
the U.S. EPA's current policy to  treat and/or minimize waste on-
site. In this removal action, MAECORP and the U.S. EPA treated
over 9,000,000 gal of contaminated water effectively and econom-
ically.

INTRODUCTION
  With the Superfund Program reauthorization, the U.S. EPA and
its contractors have a respectable challenge ahead of them. Due
to new regulations and an increase in waste volume, off-site dis-
posal is becoming  more  costly and less available.
  To meet the challenges the U.S. EPA is faced with, on-site treat-
ment is the likely alternative. On-site treatment is technically viable
and, in most cases, less  costly than off-site disposal. New tech-
nologies and existing technologies are being adapted for use on
hazardous waste sites. The systems are designed to be mobile and
meet strict discharge requirements.

                         CEOAH POINT ROAD
                                            ^^G3Q Q x
                            \FINAL OIL LAGOON
                                        PROCESSING AREA
                      T OIL SKimiNG LAGOON
            D
            ooQ
                                        WATER POND A
                          Figure 1.
      Commercial Oil Services, Incorporated, Oregon, Ohio
  The Commercial Oil Services site, with greater than 9,000,000
gal of contaminated water, was a prime on-site treatment choice.

HISTORY OF THE SITE
  Commercial Oil Services is located  in an industrial area of
Oregon, Ohio. The facility covering approximately 21 acres has
7 lagoons containing over 9,000,000 gal of water and oil (Figure
1). The site is bordered to the north by Cedar Point Road and a
Sohio refinery, to the south by Gradel Landfill (a closed sanitary
and industrial facility), to the east by an unnamed creek, and to
the west by Otter Creek Road.

Commercial Oil collected, reclaimed and re-refined waste oils from
1945 to 1985. In February, 1984, the Ohio Environmental Pro-
tection Agency performed a TSCA inspection of the facility for
PCB disposal. PCBs were found in several lagoons with concen-
trations greater than 50 ppm. In March,  1985, a PCB compliance
inspection by U.S. EPA personnel again found PCBs over 50 ppm
in several lagoons. Commercial Oil Services ceased operations in
December, 1985, due to a decline in business.  The U.S. Justice
Department filed suit against Commercial Oil for improper PCB
storage, marking and disposal in late 1985 and early 1986.
  On June 5, 1986, reports of overflowing lagoons and continuing
heavy rains prompted U.S. EPA emergency action. The action was
initiated due  to the possibility of off-site migration of contami-
nated materials. Under the initial action, water was transferred to
sequential lagoons to give the maximum freeboard in the lagoons,
thus minimizing the threat of off-site migration. Under that same
action, a small treatment system was used to gain additional free-
board. The system was deemed inadequate for the treatment
needed. Modifications were suggested but not implemented due
to a lack of funds.
  On Oct.  3, 1986, heavy rains again caused overflow of several
lagoons. Oil and water were picked up and contaminated soils were
scraped and stored on site. After the site was stabilized, crews were
demobilized until additional funding could be secured.
  On Dec. 2, 1986 MAECORP INcorporated  and was awarded
the Delivery Order to design, fabricate, install and operate a Mobile
Wastewater Treatment System (MWWTS)  at Commercial Oil
Services.

THE MOBILE WASTEWATER TREATMENT
SYSTEM (MWWTS)
  Upon issuance of the Delivery Order for the Commercial Oil
Services site,  MAECORP was tasked with designing, fabricating,
installing and operating a MWWTS that could effectively treat dis-
solved oil and grease (> 1,000 mg/1), various volatile organic com-
                                                                                                  TREATMENT     169

-------
pounds (1 to 6 mg/1), phenols (1 to 10 mg/1), heavy metals (10
to 100 /ig/1,  and suspended solids (> 100 mg/1). The system
incorporated oil/water separation, metal precipitation, neutrali-
zation, clarification, sand filtration and granular activated carbon
adsorption. In order to discharge the effluent, the treatment sys-
tem had to meet the facility's NPDES permit limitations. In addi-
tion, the system was designed to meet the more stringent State of
Ohio Water Quality Standards.
  The main portion of the treatment  system  was housed in  an
existing structure on-site. One reaction chamber, three clarifiers,
a multi-media filter, sump basins, associated pumps and a mixing
tank  were snugly fit  into a 30-ft by 60-ft area. An additional
temporary structure was built adjacent  to the existing structure to
house one adsorption cell, three granular activated carbon cells
and  an intricate flow directing manifold system. The oil/water
separator was positioned directly outside the building (Figure 2).
                          Figure 2.


   Due to the inclement weather, the  entire treatment  system
required winterization which included enclosing the entire system
in heated structures, insulating the outside influent line and running
a heat tracer along the influent pipe.
   The contaminated water was pumped approximately 1,000 ft
from the storage ponds to the treatment system which was designed
to operate at 200 gal/min. The first step of the treatment procedure
was physical separation of the aqueous  insoluble oils layer in an
oil/water separator. The water layer was pumped to the reaction
chamber and the oil layer was returned to the oil lagoon.
   In the reaction chamber, which was equipped with air lines for
mixing, 50% caustic soda was metered in to raise the pH to 10.5
to 11.0. Halfway through the chamber, alum was metered in to
accelerate flocculation of suspended and dissolved heavy metals
(lead, chromium and copper).  The reaction chamber was mani-
folded to three clarifiers in parallel.
  The original  system design predicted solids would settle as a
hydroxide sludge. After startup of the system, however, it was dis-
covered that high levels of soluble oil in the wastewater. coupled
with the vigorous mixing and chemical additions, produced a
floating material as well as settleable solids. Some modifications
to clarifiers were necessary to remove both the floating material
and settleable solids. The clarified wastewater  was pumped to
another mix tank where the  pH  was adjusted with sulfuric acid
to 6.5 to 7.0.
  The neutralized wastewater was fed through a multi-media vessel
(sand filter) to remove the fines.  The filtrate was passed through
a multi-media cell containing an absorption medium  capable of
removing 50%  dissolved oil  and grease by weight.
  The wastewater was then introduced to  three granulated  acti-
vated carbon (GAC) columns  in series. The GAC columns removed
any remaining oil and grease and removed the other organics in
the stream. The GAC cells were manifolded  so servicing could take
place without shutting the treatment system down. The effluent
was discharged directly  to an unnamed creek west of the site.
  The operation of the treatment system was complicated by two
situations; the inclement weather and respiratory protection require-
ments in the treatment building.
  The inclement weather created a number  of problems. First and
most critical, was personnel safety. Persons working outside were
subject to exposure problems (i.e., frostbite). To  prevent this, fre-
quent  breaks required additional crew size to meet the system's
operation needs.  The extreme cold also caused line  and pump
freeze.up problems. If for any reason the treatment process was
stopped, quick draining of  the  lines and pumps was  of great
importance. If any of the lines or pumps froze, the amount of time
to thaw them  would cause a significant  delay in the project
completion. Standard operating procedures for system shut-down
were  implemented to  minimize the  possibility of  equipment
freeze-up.
  Twice during the project, the conditions were so extremely cold
that the system actually froze up while in operation. The total loss
of time due to  the extreme conditions was 6 days.
  The second complication involved respiratory  protection in the
treatment building. Air monitoring at the initial startup of the
system indicated level B protection was required. With modifica-
tion to the mix tanks and an elaborate air monitoring program,
the level of protection  was lowered to level C.

                           Table 1
       Laboratory  Analysis Parameters and Discharge Limit*
Parameter
Benzene
1,1.1-TCE
Toluene
Trans -1,2- Dlchloroe Chene
Xylene
1,1-Dlchloroe Chene
Naphthalene
Phenol
Copper
Lead
Chromium (TTL)
PCB
NH3-N
Dissolved Oxygen
Oil and Grease
TSS
Effluent
Llnltatlon
5 ug/1
200 ug/1
2 mg/1
6 ug/1
440 ug/1
7 ug/1

3,500 ug/1
23 ug/1
40 ug/1
54 ug/1
<.001 ug/1
13 mg/1
4 mg/1
10 mg/1
30 mg/1
170    TREATMENT

-------
ANALYTICAL
  To monitor the treatment system's efficiencies and to ensure the
effluent was within the permit requirements, MAECORP mobilized
a comprehensive field analytical laboratory. The laboratory was
capable of performing analyses in accordance with U.S.  EPA-
approved analytical methodology and QA/QC guidelines.
  The  instrumentation in the laboratory consisted of two dual-
channel gas chromatographs with a head space analyzer, an atomic
absorption spectrophotometer with a graphite furnace, and all other
miscellaneous equipment required  for sample preparation. The
parameters analyzed and effluent limitations are shown in Table 1.
  Initially, the treatment system was monitored daily at several
points  including influent water, between carbon  cells and  the
effluent. The laboratory was staffed 24 hr/day for several weeks
until MAECORP and the U.S. EPA were satisfied that the sys-
tem operation met all discharge requirements. After it  was  deter-
mined that the MWWTS was operating efficiently, the number of
analyses was reduced significantly.
COST ANALYSIS
  Beyond limiting liabilities of off-site disposal, the economics of
on-site treatment can offer a very attractive alternative. At the initial
phase of this project, off-site transportation and disposal options
were explored. The most cost-effective transportation and disposal
option was identified at $0.25/gal. To transport and dispose of
the 9,000,000 gal of wastewater would have cost $2.25 million. This
cost does not include any site administration or operations costs.
The on-site treatment system accomplished the same quantity of
treatment which did  not include the liability associated with trans-
portation for $0.12/gal. The on-site costs were half those of tradi-
tional off-site transportation and disposal costs.

CONCLUSION
  To meet the challenges of the reauthorized Superfund program,
the U.S. EPA  and its contractors must utilize effective and
economical approaches to site cleanups.
  The U.S. EPA and MAECORP effectively and economically
treated over 9,000,000 gal of water contaminated with volatile
organics, phenols, oils  and heavy  metals.

ACKNOWLEDGEMENT
  The authors would like to express their gratitude to Elise Allen
(Weston TAT) and Karl Yost (MAECORP Incorporated) for their
invaluable assistance on the project. Thanks is also directed to the
entire operations, technical and support staff who helped make
the project a success.
                                                                                                        TREATMENT     171

-------
                 Optimization  of Remedial  Treatment  Actions  for
                                           Contaminated  Soils

                                                Edwin F. Barth  P.E.
                                     U.  S. Environmental Protection  Agency
                                    Hazardous Waste Engineering Laboratory
                                                   Cincinnati, Ohio
                                                    John  J.  Barich
                                     U.  S. Environmental Protection  Agency
                                                       Region X
                                                 Seattle,  Washington
ABSTRACT
  A conceptual process for optimizing remedial treatment actions
for contaminated soils at uncontrolled hazardous waste sites is dis-
cussed. This process involves generating a breakpoint or knee-of-
curve analysis for the many soil/waste and site characteristics that
should be involved in a cleanup decision.

INTRODUCTION
  Response actions at an uncontrolled hazardous waste site may
be separated into  operable units, allowing the  phasing  of site
response actions. Many of these operable units involve contami-
nated soils. The amount of soil involved in the  remedial action
cleanup decision (above an action level) normally is tied into the
groundwater and surface water cleanup goals and/or direct con-
tact threat at the site.
  SARA suggests treatment of Superfund site waste by decreas-
ing the mobility,  toxicity or  volume of hazardous material.
However,  treatment technologies are not easily selected and
implemented at a  hazardous waste site because of the many
soil/waste characteristics and site-specific factors that should be
involved in the site cleanup decision. Personnel involved with these
decisions should evaluate more than the chemical constituents of
the waste matrix.
  Cost is an important non-technical factor in a cleanup decision.
Since unit cost data can vary greatly among treatment technolo-
gies, consideration should be given to meet the appropriate cleanup
goals at a site by using a mixture of technologies. One technology
might be used for destroying the most mobile and toxic waste while
another technology could be used for lesser mobile and toxic waste.

OPTIMIZATION  PROCESS DEVELOPMENT
  In several past RODs for Superfund sites, only one technology
was proposed for treating the entire volume of contaminated soils.
This decision may have been based on only one soil/waste charac-
teristic such as the British Thermal Unit (Btu) value resulting in
the incineration of a large volume of soils. An optimization  process
would involve the evaluation of several soil/waste and site charac-
teristics which might indicate that several technologies may  be used
separately in the treatment of a contaminated waste medium. The
concept of optimization may have particular application  to sites
that have a mixture of  low volume, "highly" contaminated soils
and high volume,  "low" contaminated  soils.
  The decision process for implementing a treatment technology
becomes permutated when considering more than one characteris-
tic. Each applicable technology differs in treatment ability, residuals
quality, preprocessing  requirements, cost,  implementation time,
ambient conditions and area requirements. For example, inciner-
ation would be a likely preliminary choice for contaminated soils
with a high Btu value underlying an abandoned lagoon; however,
upon further site evaluation, the required depth of excavation (due
to contamination) may be too great for normal excavation prac-
tice or too deep for an effective insitu process. Furthermore, the
soils might need to be processed (e.g., finely ground) before entering
the incineration unit. Upon detailed evaluation, an optimal process
may be to remove and incinerate the top layer of soils (higher Btu),
remove and solidify/stabilize the middle layer of the lagoon (lower
Btu) and insitu solidify/stabilize the bottom layer of material.
  The concept of optimization involves generating knee-of-curve
analyses for the major soil/waste involved in the treatment decision.
For the optimization process to be useful, each soil/waste and site
characteristic  needs to have general numeric ranges related to its
degree of hazardous characteristics or cost-effectiveness. A graphi-
cal representation of the data is necessary. The ordinate normally
would be volume of soil and the abscissa would cover the numeric
ranges of a soil/waste or site characteristic. Figure 1 is an example
of a knee-of-curve analysis generated for the Fields Brook, Ohio
site.' illustrating mobility as a soil/waste characteristic. Incinera-
tion was chosen as the treatment technology for soils contaminated

                         Table 1
           Wistt/Soll Characteristics  for Consideration
                   In Optimization  Process
 mobility

 toxtcity

 partitioning coefficient

 Hydrolysis constant

 ionijation constant

 persistence

 Henry's constant

 BTU

 particle sue distribution

 viscosity

 boiling point/tlasn point

 cation exchange capacity

 shrink/swell capacity
i moisture

". radioactivity

1 solids

I chlorine

X sulfer

1 silt

l sand

t clay

I volume reduction (potential)
172    TREATMENT

-------
                                                              Figure 1
                                                   Sediment Volumes (10~6 Sediment
                                               Ingestion Risk or Greater) vs. Mobility
with a soil-water partitioning coefficient greater than 2400 ml/g
(slightly mobile).
  A detailed optimization process should consider more than one
soil/waste or site characteristic using a constant ordinate scale so
that curves can be compared or overlapped. A contradiction for
treatment technology selection might occur when curves are over-
lapped. For example, the soil contaminated with a high Btu waste
may be the  least mobile or least toxic waste on the site.
  Table 1 lists several waste/soil characteristics that might be used
during the optimization evaluation for a site.
  The major  site characteristics that  should be  included in an
optimization process include: depth of contamination, amount of
overburden, permeability, depth to water table and site area.
  These characteristics can be determined during the remedial in-
vestigation/feasibility study process. The optimization process
should  consider implementation time  as a  construction factor,
resource recovery and resource abandonment.
  An example of a site/waste characteristic that may be used in
an optimization process is particle size. At the Minker, Missouri
hazardous waste site, 69% of the total mass of dioxin present in
the soil was associated with soil particles with a size of less than
53 an2. Remedial action  could be optimized if a  pretreatment
process could separate these particles from the remaining soil, there-
by allowing the option  of two different treatment technologies,
assuming their technologies treatment efficiency depended on soil
size. Another example of optimization would be the separation of
gravel size particles from other contaminated soils. The gravel may
be able to be separated, then washed with water or other extrac-
tion fluid, resulting in a reduced soil volume for treatment by
another process.  The rinsed water may be useful  in processes
requiring water such as solidification/stabilization.

CONCLUSION
  An optimization process involving the consideration of several
soil/waste and site characteristics for  treatment of contaminated
soils at an uncontrolled hazardous waste site may be useful in a
cost-effective decision process. As "acceptable" numeric ranges
for soil/waste and site characteristics are not available, it may be
difficult to implement for a site decision. Further guidance for
optimizing remedial actions continues to be developed.

REFERENCES
1. U.S. EPA, "Superfund Record of Decision: Fields Brook Sediment
   OH," Number EPA/ROD/R05-86/035, Sept. 1986.
2. Vick, H.W., et al., "Physical Stabilization Techniques for Mitigation
   of Environmental Pollution from Dioxin Contaminated Soils," Science
   Applications International Corporation (publication pending).
                                                                                                           TREATMENT     173

-------
                     Chemical  Oxidation  Destruction  of  Organic
                                 Contaminants  in  Groundwater
                                                  Donald G. Hager
                                               Rubel and Hager, Inc.
                                                  Tucson, Arizona
                                                   Carl G. Loven
                                               Christopher L. Giggy
                                            Peroxidation Systems, Inc.
                                                  Tucson, Arizona
ABSTRACT
  Extensive field testing of the ultraviolet light catalyzed hydro-
gen peroxide (UV/H202) process has demonstrated the applicability
of this emerging new technology to a wide range of contaminated
groundwater problems. Pilot  plant size  equipment has been
employed to treat contaminated water from numerous locations
throughout the United States to obtain performance and cost data
on the destruction of solvents, pesticides and fuels found in con-
taminated  potable water aquifers and  land-fill  leachates. The
concentration of organics in the untreated water has ranged from
less than 1 mg/1 to 9,000 mg/1.  Application of the UV/H202
process produced treated water with non-detectable concentrations
of the contaminants.
  Design development data,  as  well as projected capital and
operating costs,  are presented for four groundwater purification
systems. Four systems have been  designed; two have progressed
into full scale operation.
  The  UV/H202  process is  a cost-effective,  on-site  organic
destruction technology. Organic contaminants, in either dilute or
concentrated  aqueous  mixtures, can  be  destroyed  without
deleterious air emissions or  production of residual waste  by-
products.
INTRODUCTION
  In 1982 a full scale groundwater purification system employing
a UV/H202 process was installed in Colorado to treat  150,000
gal/day of water contaminated primarily  with tetrahydrofuran
solvent. This installation has been  highly successful in  the destruc-
tion of various solvents found in the water.  This UV/H2Oj process
employs technology which was available at the time and  features
low energy UV power and an open-to-atmosphere reactor.
  In September 1987, two UV/H202 (Perox-pure ™) groundwater
systems were installed in California greatly  advancing the state of
the art of chemical oxidation of organic contaminants in water.
Both the treatment systems, 600,000 gal/day and 60,000 gal/day,
respectively, employ proprietary high energy UV sources and com-
pletely sealed  reactors. These advanced process design  features
reduce the oxidation  time, reduce  the size  of the reactor and
eliminate the possibility of fugitive air emissions.
  These systems are designed to reduce the concentration of organic
solvents including methylene chloride, chloroform, vinyl chloride,
toluene, trichloroethylene (TCE), 1,2-dichloroethane (DCA), trans
1,2-dichloroethylene (DCE) and benzene to non-detectable levels
of concentration. On-site  field studies of  this process provided
process data needed for the design of the  full scale  systems.
  These state of the art Perox-pure™  installations and others
under design confirm the practical cost-effectiveness of chemical
oxidation to many organic contaminantion problems in ground-
water and industrial effluents. The organic chemicals  listed in
Table 1 have been successfully treated by the described process and
illustrate  the applicability of  treatment over a broad range  of
contaminants.
                     Table 1
            Organic ConUunhuols
          Amenibk to L"V/H2O,
 Acetic Acid
 Acetone
 Acetonitrile
 Aldecarb
 BOD
 Benzene <
 Butyric Acid
 Carbon Tetrachloride
 Chloroacetic Acid
 Chlorobenzene
 Chloroform
 Chloromethane
 COD
 Color
 Chlorophenol
 Cyanide Compound!
 Cyclohexane
 Cyclohexanone
 Dibromodichloropropane
 Dibutylphthalat*
 Dichlorobenzene
 Dlchlorocthane
 Dichloroethylene
 Dichloropentadiene
 Dichlorophenol
 Dichlorophenoxyacetic Acid
 Dichlorotri fluoroethane
 DimethyIbutane
 Dithane
 Ethanol
 Ethylbeniene
 Ethylcyclobutane
 Ethylenediaminetetraacetic Acid
 Ethylenedibromide
 in Witcr
Treatment
      Formaldehyde
      Tormic Acid
      Treon-TF
      Hexachlorobutadiene
      Bexachloroethane
      Hydrazine
      Xsopropanol
      Methylcyclopentane
      Methylpentane
      Methylene chloride
      He thylethyIketone
      MethyliiobutyIketone
      Moropholine
      N-Butylamine
      Napthalene
      Pentachlorophenol
      Pentane
      Phenol
      Propionic Acid
      Resorcinol
      Sodium TMocyanate
      Sulfolane
      Tetrachloroethane
      Tetrachloroethylene
      Tetrachlorophenol
      Tetrahydrofuran
      TOC
      Toluene
      Trichlorobenxene
      Trichloroethane
      Trichloroethylene
      Trichlorophanol
      Vinyl Chloride
      Xylene
174    TREATMENT

-------
  The UV/H2O2 Process  converts  organic  contaminants to
carbon dioxide and water without creating air emissions or residual
waste streams. The on-going development of this on-site, in-situ
destruction technology has been described in previous publications
'•2i3. Process operating variables, water chemistry and equipment
design all are important to the successful destruction of hazardous
and toxic  contaminants.

DESIGN  DEVELOPMENT STUDIES
Testing Equipment
  The groundwater from four contamination sites was treated in
a series of tests using a pilot scale UV/H2O2 chemical oxidation
system '•2-3 mounted in a Field Test Mobile Unit.  The chemical
oxidation pilot system consisted of three interchangeable stainless
steel  cylindrical reactors, a variable speed pump, a hydrogen
peroxide  (H202) feed pump and a stainless steel feed reservoir
equipped  with a cooling coil  for temperature control. A high
intensity  ultraviolet  (UV) lamp enclosed in a quartz tube was
installed axially in the reactor. The pH and temperature of the
solution were monitored throughout the reaction. All materials in
contact with the test solution were stainless steel, viton or quartz.

Testing Procedure
  The purposes of the on-site groundwater testing program was
to assure the destruction of all organic contaminants to California
Department of Health Services Regulatory Action  Levels and to
perform  a comprehensive  optimization of process  conditions.
UV/H202 testing variables included UV dose,  H202 dose, oxida-
tion time, catalyst concentration and pH. The concentration of
catalyst, when used, was very low in order to avoid the need to
remove   residual contaminants  remaining  in  the  treated
groundwater.
  Data collected during each test run were used to establish the
parameters for subsequent tests. Each test was designed to measure
the effect of a key process  variable on the oxidation rate of the
organics in the groundwater.
  The untreated and treated water samples were analyzed by Gas
Chromatography/Mass  Spectrometry.

Groundwater  Site A
  The Site A location  employs  a series  of 12 extraction  wells
manifolded to  a single treatment site. The extracted water is purified
by  the UV/H2O2 Process and is reinjected into the ground. The
groundwater contamination originates from leaking solvent and
fuel tanks and manufacturing process areas which use organic
chemicals.
                                                                             800
           600
                       O Tetrachloroethylene
                       • 1,1-Dichloroethylene
                       A Trichloroethylene
         cr-
         4J
         id
         C
         0)
         o
         c
         o
         u
           400
200-*
                        1      2        3
                     Treatment Time,  Min.

                              Site  A

                           Figure 2
  Testing  Results—The optimum conditions  for  treating the
groundwater of Site A were established as follows:
UV Dose:
H202 Dose:
pH:
Oxidation Time:
160 watts/1
7 mg/l/min
7.5
2.5 min
  Table 2 contains the purification results for Site A groundwater
which was treated at optimum operating conditions. The primary
contaminants, tetrachloroethylene and 1,1-dichloroethylene, were
not detected after 2.5 min of treatment. Trichloroethylene also was
completely destroyed in that period, but a residual amount of the
intermediate  oxidation  product,  1,1,1-trichloroethane, still
remained.  Very little Freon-TF was removed under these condi-
tions; however, the starting Freon-TF concentration was consider-
able  lower than the Regulatory Action Level of 18 mg/1.
                                                                                            Table 2
                                                                      Site A Groundwater Treatment Data Effluent Concentrations
                                                                               Given as a Function of Treatment Time
Treatment Time
(min. )
Contaminant
Tetrachloroethylene
1 , 1-Dichloroethylene
Freon-TF
Trichloroethylene
1,1, 1-Trichloroethane
0
704
263
71
54
2.5
ND
ND
75
ND
29
5
ND
ND
53
ND
20
Regulatory t
Action Level
4
6
18,000
5
200
                          Figure 1
                                                                    California  Department of Health Services Regulatory
                                                                   Action Levels  for contaminants  in drinking water.
                                                                  Groundwater Site B
                                                                    The groundwater Site B location employs a single groundwater
                                                                  extraction well which supplies influent to the UV/H2O2 treatment
                                                                  system.  The  organic contaminants  found in  the  groundwater
                                                                  originated from the use of cleaning solvents at this manufacturing
                                                                  site.
                                                                                                        TREATMENT     175

-------
Testing Results—The optimum treatment conditions, established
for Site B, were as follows:

  UV Dose:
  H202 Dose:
                   160 watts/1
                   10 mg/l/min
  pH:              7.1
  Oxidation Time:   4 min

  Table 3 summarizes data obtained with optimum treatment con-
ditions for  Site  B  groundwater.  100%  of the  trans
1,2-dichloroethylene and 91%  of  the trichloroethylene  were
destroyed in 2 min.  Although the trichloroethylene residual con-
centrations in the 2 min sample did not meet the prescribed Regula-
tory Action Level, no trichloroethylene was detected after 4 min
of treatment. Figure 3 presents the data for Site B groundwater
treatment in graphic form.
                         Table 3
              Site B Groundwater Treatment Data
               Effluent Concentration Given as a
                 Function of Treatment Time
                 Treatment Time (min.)
 Contaminant
                                            Regulatory   ,
                                            Action Level
 Trichloroethylene           1924  57  ND         5
 trans  1,2-Dichloroethylene   198  ND  ND        16

  California  Department  of  Health  Services  Regulatory
 Action Levels  for contaminants in drinking water.
                                                                         16,000
                                                                     rH   12,000
                                                                      C
                                                                      o
                                                                     •H
                                                                     4J
                                                                      4J
                                                                      e
                                                                      0)
                                                                      o
                                                                      o
                                                                      o
    O  1,2-Dichloroethane
    •  Chloroform
    A  Methylene  chloride
                                                                                                 T
                                                                                                 30
    15     30      45     60

Treatment Time,  "Sin.

         Site  C


   Figure 4
                                                                                        Table 4
                                                                            Contaminant Concentration Ranges
                                                                                       for Site C
                                                                 Contaminant
                                                                                               Concentration (uq/1)
           2500 J
                      O Trichloroethylene
                      • Trans  1,2-
                               Dichloroethylene
                         123
                     Treatment Time,  Min.
                              Site  B


                         Figure 3
 Groundwater Site C
 Testing Results—The concentrations of the major contaminants,
 1,2-dichloroethane, chloroform and methylene chloride, varied
 somewhat from sample to sample. The range of concentrations
 of these contaminants are shown in Table 4.
                                                                 Chloroform
                                                                 1,2-Dichloroethane
                                                                 Methylene chloride
                                                                                                   8,200  i  3,400
                                                                                                  16,000  1  6,800
                                                                                                   2,060  i  860
                                                                 The following optimum treatment conditions developed for Site
                                                               C groundwater were as follows:

                                                                 UV Dose:        160 watts/1
                                                                 H202 Dose:       150 mg/l/min
                                                                 pH:             8.1
                                                                 Oxidation Time:  45 Min

                                                                 Table  5  shows the test results for Site C groundwater. The
                                                               1,2-dichloroethane and methylene chloride were rapidly destroyed,
                                                               as were the secondary contaminants present in the 15 min. Chloro-
                                                               form, however, was more resistant to destruction. After approxi-
                                                               mately 40 min, the chloroform concentration was reduced to below
                                                               The Regulatory Action Level. The data reflecting destruction of
                                                               organic contaminants from Site  C groundwater are presented
                                                               graphically in  Figure 4.
                                                                                          Table 5
                                                                              Site C Groundwater Treatment Data
                                                                   Effluent Concentration Given as a Function of Treatment Time
                                                                                    Treatment Time (min)
Contaminant
Chloroform 7
1, 2-Dichloroethane 16
Methylene chloride 3
Carbon tetrachloride
1,1, 1-Trichloroethane
0
,900
,900
,000
	

15
2,300
110
230
45
56
30
330
ND
ND
ND
ND
45
38
ND
ND
ND
ND
Regulatory .
60 Action Level
3
ND
ND
ND
ND
100
1
40
5
200
                                                                  California Department of Health Services Regulatory
                                                                  Action Level* for contaminants in drinking water.
 176    TREATMENT

-------
Groundwater Site D

  The  Site D groundwater extraction well  is located near an
abandoned unlined lagoon in which  a variety of wastes from
manufacturing processes were dumped over  a period of time.
  Testing Results The most favorable treatment conditions for Site
D groundwater purification were as follows:

UV Dose:         590 watts/1

H202 Dose:        300 mg/l/min

pH:               7.6

Oxidation Time:   45 min

  The chemical oxidation treatment data for Site D are summarized
in Table 6 and depicted graphically in Figures 5-8.
                          Table 6
               Site D Groundwater Treatment Data
   Effluent Concentrations Given as a Function of Treatment Time
                    Treatment Time (min)
       o
       •H
       4-1
       (0
       M
       4J
       c
       0)
       O
       c
       o
       CJ
           800_
           600
400-
           200
                      Freon-TF
                      1,1,1-Trichloroethane
                      1,1-Dichloroethylene
Contaminant
Methylene chloride 4
Chloroform 1
Carbon tetrachloride 1
Freon-TF
1,1, 1-Trichloroethane
1 , 1-Dichloroethylene
Trichloroethylene
Toluene
Tetrachloroethy.lene
1 , 1-Dichloroethane
Vinyl chloride
0
,254,333
,843,016
,216,775
674,905
194,245
166,591
74,844
48,792
29,068
19,523
15,486
30
3,600
2,500
390
ND
820
ND
ND
ND
ND
ND
4.4
Regulatory t
Action Level
40
100
5
18,000
200
6
5
100
4
20
2
                                                                                       Treatment Time,  min.

                                                                                              Site  D


                                                                                         Figure 6
           5,000
                        Methylene chloride
                        Chloroform
                        Carbon Tetrachloride
conditions. The proprietary engineering technology responsible for these
scale-up results is subject to pending patents which precludes detailed dis-

cussion at this time.
  Table 7 summarizes these scale-up design criteria when converting pilot
data to full scale equipment for Sites A-D.
                                                                                75
                                                                                           Trichloroethylene
                                                                                           Toluene
                                                                                           Tetrachloroethylene
                                       30
                      Treatment  Time, min.

                              Site D



                         Figure 5
 DESIGN CRITERA AND PROJECTED COSTS

   Much higher efficiency of oxidation has been demonstrated in full scale
 systems equipment compared to the standard pilot scale testing. Depending
 on the concentration and type of contaminant in the particular water, as
 well as the effluent objective for the treatment process, the net effect has
 been a 2- to 4-fold decrease in oxidation tune using the same basic operating
                                                                           Cr>
          O
          •H
          4J
          al
          M
          4J
          C
          
-------
              25
                         1,1-Dichlorethylene
                         Vinyl Chloride
         o
         0)
         O
         o
         u
                      Treatment Time,  nun.
                             Site  D
                         Figure 8
                           Tablet
             Projected Coil* of UV/HzOj Treatment
                                  $  per 1000 Gallone
 Operating Coats

 Energy * S0.075/KWH

 H202 * $0.7S/pound

 Maintenance  t  5% of
  Capital Costs
                 •
 Operating Labor

 Amortization of Capital
  9 20* per year

 Total
Site A  Site  B  Site C  Site D

 $1.12   $1.41  $11.25  $15.00

  0.04    0.07    1.10   33.OS
  0.04


  None

  0.17
0.12


None

0.48
0.57


None

2.28
2.09


None

8.37
                                                                                            $1.37    $2.08  $15.20   $58.51
                                                                  Only periodic surveiHence  is  required
to range from $1.37/1000 gal for low levels of concentration to
$58.51/1000 gal for highly contaminated leechates. Table 9 relates
costs of treatment with flow volume and contaminant levels in the
groundwater.
  Based upon these design development data, full scale chemical
oxidation systems have been designed for Site A and Site B. The
UV/HjOj equipment supplied  for Site  A  is shown in  the
photograph.
                           Table 7
                   Scale-up of Testing Data to
                Full Scale Equipment for Sites A-D
Testing Data
Site
A
B
C
D
Full
A
B
C
D
UV Dose
Watts/Liter
160
160
160
590
Scale Design
200
200
200
200
Oxidation
Time (min.)
2.5
3.0
30
45
Criteria
1
1.2
12
45
H202 Dose
(mg/1)
17.5
30.0
600
13,500
6.9
11.8
177
6750
                                                                                           TaMe 9
                                                                            Summary of Process Criteria and Costs for
                                                                         Site A-D GrooBdwaler PwUkatioB with UV/H,O,
                                                                                                                  Caplul *

Sic.
A

I

C


D






riov
Iซcซ Principal
(ปpป) CoaUBliuati
400 TCI
DCI
40 TCI
DCI
15 CF
OCA
NCI
15 HCI
CF
CC1
frซ
-------
                                     Mobile Waste  Oil  Recovery

                                                    Craig A. Nowell
                                                    Process Engineer
                                                     Mark J. Hardy
                                                    Systems Engineer
                                          Bird Environmental Systems, Inc.
                                            South Walpole, Massachusetts
ABSTRACT
  New federal regulations regarding the disposal and cleanup of
oily wastes have generated renewed interest in mobile oil recovery
systems. Bird Machine Company, Inc., of South Walpole, Massa-
chusetts, formed  a new  company called Bird Environmental
Systems, Inc., to design, manufacture and provide field process
services using the combined Bird separation and waste oil process
technology. With the technology mounted on a self-contained 48
ft. trailer, a complete system can be mobilized at the generator's
site. This mobile system can handle a variety of oily wastes and
can transform a waste oil stream into three streams that are either
reusable or easily disposed.

INTRODUCTION
  Oily waste found in today's refineries and production fields is
a mixture of many types of oily wastes consisting of tank bottoms,
API separator sludges,  cat,  fines, etc., that cannot be separated
by  today's  single  gravity  conventional  chemical  treatment
technology.
  The wastes that have been generated and stored over a period
of time have separated  under normal gravity leaving an oil and
water that can be decanted The resultant waste left in the tank is
a middle emulsion layer. After a period of time the emulsion layer
builds up in the tank. At this time emulsion "breaking" chemicals
are added. These chemicals usually produce free oil and water layers
that can be decanted leaving a mixture of emulsion and chemicals.
Drawing off the decanted liquids provides more volume so that
the generator can add more waste to  the available volume.
  This cycle continues until there is no more available volume to
store waste and the chemicals no longer work. At this time, the
tanks  or pits are full of a mixture of oily wastes (emulsions) and
chemicals that no longer can be treated and recovered  through
gravity separation.
  The unfortunate problem with these tanks and pits, in addition
to the variety of chemicals present, is the  lack of uniformity in
the distribution  of oil, water and solids. The top of the tank, in
most cases, will have more oil; the opposite is true for the bottom
where more water and solids are present. Another problem is that
most tanks and pits used for storage and accumulation are often
too large for proper mixing to result from conventional agitation
or recirculation. This results in a varying oil-water mixture and
a range of  chemical  compositions that hampers most recovery
systems. This lack of uniformity and constantly changing chemical
composition requires a versatile oil recovery system that combines
for both chemical treatment and mechanical separation.
  The chemical treatment system may involve the addition of
emulsion breaking chemicals, oxidizing agents to control hydrogen
sulfide, polymer to assist in water clarification and pH adjustment
chemicals to help both control hydrogen sulfides and water wet
the solids. Thus, the goal is to assemble an oil recovery system that
provides the ability to handle the variable process streams and to
perform the oil recovery process economically.
  Before a truly good mobile oil recovery system can be designed
and built, a brief background on the chemistry of emulsions and
how to break them must be explored. This background informa-
tion will give insight into the complexity and flexibility of the mobile
system and why the two must be combined to provide a system
that is cost-effective and possesses the capability to handle a variety
of waste streams.

OIL RECOVERY
Theory  of Emulsions
  Refinery emulsions can be defined, generally, as an intimate two
phase mixture of oil and water with one phase dispersed as droplets
in the other phase.
  All emulsions are stabilized by an interfacial film coating of the
discontinuous phase. This stabilizing film can consist of a com-
plete mixture of dissolved and colloidal matter and  suspended
solids. This film also may exhibit the following characteristics:

• High  viscosity
• Gelatinous form due to a variety of waxes
• Electric charge on the particles

  Generally, an emulsion's stability is not a function of the degree
of dispersion; rather, it is a function of the stabilization effects
of viscosity, surfactants, electrolytes and suspended solids. This
is illustrated by the following five categories that affect the stability
of an emulsion:

• Strength  of the interfacial surface tension
• Orientation of the emulsifying agent and the solids particles in
  the emulsion
• Viscosity of both the continuous phase and the emulsion itself
• Type  and amount of emulsifying agents present
• Physical  conditions that created the emulsion

Emulsion Breaking  Theory
  The first step to a successful oil recovery system is the selection
of the proper pretreatment system. The initial process in the treat-
ment system is heating. Most emulsions can be treated and broken
at temperatures of 150 to 200F. The applied heat has several effects
on the emulsion:
                                                                                                      TREATMENT    179

-------
• Viscosity of the oil phase is reduced.This viscosity reduction is
  achieved when the heat creates a phase change that disrupts the
  interfacial waxy films that tend to trap both  solids and water.
• The densities of the oil phases are reduced. This density reduc-
  tion is due to the thermal expansion coefficient of oil.  The
  reduced density of the oil phase increases the density difference
  between phases, increasing the separation efficiency of sub-
  sequent centrifugal treatment.
• The vapor pressure  of  the water increases. The increase of
  internal vapor pressure tends to rupture the film around the water
  particles.
• Heat also increases the effectiveness of both naturally occurring
  surfactants and surfactants added to the emulsions. Both types
  of surfactants lower the surface tension of the interfacial  film.

  Heat, therefore, weakens the interfacial films of various emulsi-
fying agents. Weakening then allows coalescence of the oil and
water and the removal of the solids.
  hi addition to the use of heat, the emulsion  may be dispersed
through the addition of de-emulsifiers and wetting agents.  The
wetting agents  are used to adjust the interfacial surface tension
on the surface  of solids, thus allowing the solids to be removed
from the emulsion interface. With this third phase 'solids' removed
from the interface of the  emulsion, the emulsion now becomes
much easier to break.
  Wetting agents in most cases  can either be reactive anionics
(OH- and PO4- or cationics (H + , FE++ and Al+ + +).  These
types of reagents may be referred to as pH adjustment chemicals.
This term is some what of a misnomer. The addition of acid or
caustic solutions provides a more complex form of emulsion
breaking and, in most cases, is not used merely as a pH adjust-
ment, but rather as a wetting agent. The addition of these chemi-
cals provides the following aids to emulsion breaking:

• Solubilization of components  in the interface so they can be
  removed in the miscible phase
• Promotion of ion exchange reactions to occur  affecting the solu-
  bility of the de-emulsifiers
• Promotion of precipitation reactions that help remove the emul-
  sifiers from the interface

  Demulsifiers, wetting agents,  surfactants and  heat are all
important in the scheme of emulsion breaking.
  The next step in the design of a mobile oil recovery process system
is to incorporate or integrate chemical  treatment techniques with
physical or mechanical separation processes  into a  total system.

Mechanical  Separation Technology Centrifugalion
  The advanced mobile oil recovery system  takes the  form of a
two-stage process which utilizes an advanced three phase decanter
and  a high speed three-phase self cleaning disc centrifuge.
  The advanced three-phase decanter centrifuge is designed with
the versatility to be "fine tuned" while  processing a waste oil
stream. The decanter centrifuge is a specially designed drum and
scroll (conveyor) that rotates about a  horizontal axis. The rota-
tion and physical size provide the oil recovery system with a gravita-
tional force of up to 3,000 times that of gravity and up  to a
30 sec. retention time.
  The three'phase decanter centrifuge then works with the emul-
sion breaking chemistry to  produce for a disposal or reuse a water-
wet solid, a clean water phase and an oil phase containing a small
amount of remaining emulsion.
  In some cases, process requirements dictate  that there be two
clean liquid  streams. In these cases, a three-phase disc centrifuge
must also be employed to polish one of the  liquid  streams. The
three-phase self cleaning disc centrifuge used on  the mobile system
operates at high centrifugal forces of up to 7000 times gravity and
provides a large amount of settling area in a small volume. The
disc centrifuge  in comparison to the decanter is limited to a feed
stream containing a maximum of 10<% solids and  emulsion (by
volume), making use of a disc centrifuge on the raw feed im-
practical.
  The typical results for a disc centrifuge operating on a stream
containing a small percentage of solids and rag (emulsion) would
be a clean oil stream in most cases containing less than 1 % BS&W
(bottom  sediment and water) and a clean water stream that can
be returned to a water  treatment system or combined with the
decanter centrifuge water stream.
  In summary, there are two  areas to consider in the design of
a mobile waste oil recovery system. First, one must efficiently use
of heat and chemical pretreatment to reduce the emulsion phase
of the waste, thus maximizing the amount of clean oil that can
be recovered from the waste stream. Second, one should have a
versatile  mechanical operating system to clarify the liquid phases.
This mechanical separation equipment must have the ability and
the flexibility to optimize its separation performance while in oper-
ation in  order to compensate  for varying feed conditions.

Mobile Oil Recovery System
  The Bird Environmental Systems Model 300 Mobile Oil Recovery
System was designed to meet the demands of the waste oil industry
(Figure 1). The Model 300 Oil Recovery System performs a three-
phase liquid-liquid-solid separation. The primary separation occurs
in a specially modified  Bird decanter centrifuge, and separated
liquid may be polished using a Bird Disc centrifuge. The Model
300 is designed to handle a variety of oily waste streams ranging
from 0 to 100% water, 0 to 100% oil and up to 20% solids at a
feed rate up to 100 gal/min.
                           Figure 1
                           Model 300
       TO*.
                           Figure 2
180    TREATMENT

-------
Process Overview
  The following  discussion is a general process overview of the
mobile oil recovery system. This general process schematic can be
seen in Figure 2.
  The feed rate to the process is regulated by a pair of air operated
valves which allow any portion of the feed fed to the system to
be recycled back to the source. The pneumatically controlled feed
valve system can use an operator-specified set point to automati-
cally control feed rate. This control system allows the customer
to supply any amount of feed to the system without the possibility
of upsetting the system operation. The feed pH may be adjusted
through the use of a chemical feed pump and addition tees located
at various locations along the feed line to allow optimization of
the system.
  The feed temperature is raised using a spiral tube heat exchanger
designed to heat a 100 gal/min stream of oily waste to 180 to 200F.
Separated water from the decanter centrifuge can be injected into
the feed line prior to the heat exchanger. The centrate recycle has
been included in the design to maintain design velocity to prevent
exchanger fouling and conserve on  the energy demands of the
system. Heated feed then enters a bank of hydroclones where high
density grit is removed,  thereby extending the life of all downstream
components.
   If necessary, the heated feed is chemically treated by the addition
of de-emulsifiers to enhance the capabilities of centrifugal separa-
tion. De-emulsifiers commonly used can be added in a variety of
locations in the feed line for the  added flexibility of optimizing
the contact time.
  A polyelectrolyte  flocculant is then premixed in a polymer
blending system.  The mixed polymer is sent to an aging tank to
let the polymer hydrate. The aged polymer is then fed by a posi-
tive displacement pump, diluted, mixed and fed into the centrifuge
feed line where it is mixed into the  feed stream.
   The treated waste is then fed to the decanter centrifuge where
the three phases can be separated. The solids phase is discharged
into a screw conveyor where a solidification material may be added
when necessary to decrease the free liquid content of the solids prior
to discharge into the solids receptacle. The heavy liquid phase (i.e.
water) may be either recycled to maintain heat exchanger velocity
or sent to a liquid receptacle or water treatment facility. After being
skimmed, the light phase (i.e. oil) can be sent to the disc centrifuge
if further clarification is necessary.
  After the introduction of additional emulsion breaking chemi-
cals,  if necessary, and reheating in  a  small spiral  tube heat
exchanger, the light phase effluent is sent to the  disc centrifuge
where the final clarification is performed. The liquid streams are
then discharged to the proper receptacles or discharge systems. The
solids discharge from the disc is sent to either the screw conveyor
or is fed back into the main feed line of the decanter centrifuge
for reprocessing.
  For the protection of both the operators and air quality, the
Model 300 is equipped with a vapor recovery system that draws
vapors off various containments where hazardous gases may be
generated. A centrifugal fan  collects the vapors and sends them
to two filters: (1) an activated carbon filter to remove hydrocarbons
and (2) a hydrogen sulfide filter.

Design Considerations
  The mobile oil recovery system is designed as a self-contained
separation equipment system. The system has been constructed on
a 48 ft. semi-enclosed  flat bed trailer complete with double con-
tainment features, a control room enclosure for ease of operation
and a complete lighting system for 24-hr operation and operator
safety.
  The equipment for inclusion on the mobile oil recovery system
was based on the following criteria:

• Ability to handle a varying process stream without shut down
  to make process optimization changes
• Equipment that is sturdy and has proven reliability for over the
  road travel
• Equipment and instrumentation that is both lightweight and
  compact in size, due to obvious limitations of the trailer
• Corrosion resistance to sulfide attack and most other oil field
  chemicals

  The equipment, piping supports and vibration isolation design
have been aided by computer vibrational analysis to reduce equip-
ment failure due to road and  operation induced vibration.
  To minimize corrosion, the fluid transfer system on the mobile
unit is constructed of 316 stainless steel. To minimize the risk of
leakage in the piping system,  welded fittings are utilized where
appropriate. To aid in the  decontamination of the unit  when
moving from site to site, all piping has low point drains and high
point vents to allow for flushing and complete draining.
  The protection and  safety of the environment was a  major
concern in the design criteria of the recovery system. For this rea-
son the design included:

• A vapor recovery system to  control the release of both hydro-
  gen sulfide and hydrocarbons emissions into the atmosphere
• A reagent system for addition of oxidizing additives into the feed
  stream to control dangerous  emissions of sulfides that could be
  released
• An addmix system to add solidification or stabilization addi-
  tives to the solids discharged from the system
• Double containment of all process equipment on board the unit,
  accomplished  by using an 8-in. containment wall around and
  under the equipment to contain spillage or leakage of any of the
  feed components

  The control and electrical components on the mobile system are
 explosion proof for hazardous atmospheres (Class 1 Group D
Division  1).  The  control system  utilizes  a General Electric
Programmable Logic Controller to help the operator control  the
unit. Pneumatic systems are also utilized for process control and
instrumentation and are located within the control room. The entire
control system has been designed for both the  ease of operator
control and the minimization of operator attention. These features
alone reduce the number of operators  required to  operate  the
recovery system, thus making  it more economical.
  The design and rationale for several major  pieces of equipment
selected for the oil recovery system deserve more discussion: heat
exchangers, decanter centrifuge and self-cleaning disc centrifuge.
  The heat exchangers were chosen to provide primary process
stream heating of the feed to the decanter and reheating of  the
oil stream directed to the disc  centrifuge. Both  are of the  spiral
tube design. These heat exchangers were chosen  due to their high
efficiency, and low fouling factors, and an ability to handle a high
solids content of up to 30% without plugging if the design velocity
is maintained.
  The three-phase decanter centrifuge  is specially  designed to
provide the versatility to optimize conditions  while processing the
feed stream. This controllability has been accomplished by incor-
porating an adjustable skimmer for removal of one liquid  phase
and an automatic hydraulic backdrive unit for varying the solids
removal rate. The purpose  of the hydraulic backdrive  in  the
decanter is to automatically adjust the internal conveyor speed to
insure that the optimum solids removal rate is maintained during
varying solids loading conditions of the feed stream. The adjusta-
ble skimmer will then allow for changes to be  made to the position
of one liquid phase discharge point during normal operation  and
compensate for changes in the composition in the waste oil stream.
  The three-phase disc centrifuge chosen for the mobile system
is a self-cleaning disc centrifuge designed with an infinitely variable
solids discharge timer. The operator can adjust  the liquid/liquid
interface within  the disc  stack to effect better separation. As a
result, the self-cleaning centrifuge on the mobile oil recovery system
                                                                                                         TREATMENT     181

-------
 has the versatility to compensate for changes in the waste stream
 compositions

 CONCLUSION
   In this re-emerging field of waste oil recovery, a mobile system
 that  incorporates  the  environmental, safety and operational
 concerns of the industry has been developed. This mobile tech-
 nology, the combination of Bird centrifugal separation technology
 with emulsion breaking technology, can transform refinery and

                            Table 1
                Actual Versus Projected Performance
                  Of Mobile Oil Recovery System
                                        Projected

 Run Rate;  bbl/day                            1400

 Oil Phase

   recovery efficiency, % by wt.               78
   oil,  % by wt.                              99.36
   suspended solid!, % by wt.                 0.079
   water, % by wt.                            0.560
   BS4W,  %  by wt.                             0.639

 Water Phase

   recovery efficiency, % by wt.
   water, %  by wt.                            99.7
   oil, % by wt.                              0.208
   suspended solids,  % by wt.                 0.110
   oil d solids, I by wt.                      0.318

 Solid Phase

   recovery efficiency, I  by wt.              99.9
   suspended solids,  % by wt.                 23.0
   oil, % by wt.                               33
   water, %  by wt.                             44
   cake bleed test, pass/fail                 pass

Enhancement Chemical Usage:

  emulsion  breaker,  mg/1  of  waste feed        1000
  polymer,  ปo/l  of waste  feed                 2SO
  polarity  adjustment, mg/1  of  waste feed     2000
  run temperature  ฐF                         180
Actual

 3250
 86
 99.40
 0.066
 O.S30
 0.596
 95
 99.8
 0.198
 0.020
 0.218
 99.8
 38.0
 27
 35
 pass
300
00
200
180
oil field wastes into three usable or disposable streams consisting of:

• An oil that can be mixed with the crude and reprocessed or sold
  as a fuel or a fuel additive
• A water that can be handled by the refinery's waste water treat-
  ment system
• A solid that has no free liquids and represents as little as 10%
  of the original waste volume

  The main key to the success of the mobile system approach,
though, is its cost-effectiveness. The flexibility of the advanced
decanter-disc combination and the complete system approach
provides reduced setup  time and less operator attention than many
mobile or semi-permanent skid systems. The mobile oil recovery
system shows significant cost savings over the high capital and
operational costs associated with a fixed installation.
  In actual operation the mobile system met or exceeded the per-
formance predictions  indicated in  preliminary application  and
laboratory testing. These results are listed in Table 1. This table
lists both the projected and actual performances of the mobile
system. It is  important to note that the tank processed  varied
markedly both from the sample received by our applications labora-
tory and within the tank itself. Despite these variations the system
described in this paper combined the flexibility to process the tank
and perform  within specifications.
  The mobile system, despite the up  to 50% variation in feed com-
position, recovered 0.53 bbl/min of oil. The oil exceeded most
crude specifications containing only 0.596% by weight BS&W and
if sold at current crude market prices represents a recovery value
of up to $15,300 for a full day of  operation.
  The information provided above and  in Table  1  presents
conclusive evidence that the technology and design of the mobile
oily waste process system will be an important force in the treat-
ment  and resource recovery of both current oily waste from on-
going production and the pits, ponds and lagoons from previous
operations. With this combination of cost effectiveness and flexi-
bility, a new  design of proven technology has emerged.
182    TREATMENT

-------
                        Electrochemical  Oxidation of  Hexone and
                                         Other  Organic  Wastes

                                                 Alex G. Fassbender
                                              Peter M.  Molt on, Ph.D.
                                           Pacific Northwest Laboratory
                                                Richland,  Washington
                                                   Greg Broadbent
                                         Northwest College and University
                                               Association for Science
                                                 Berkeley,  California
ABSTRACT
  Research to develop a continuous flow system for destroying
hazardous organic wastes using electrochemical oxidation is being
conducted at Pacific Northwest Laboratory (PNL). Initial batch
tests using cerium (IV) showed that a wide range of organic solvents
can be destroyed through electrochemical oxidation. Researchers
at PNL have shown that this technology can continuously destroy
pure hexone in a laboratory setting.
  Because of existing needs at the Hanford Site in southcentral
Washington State the primary focus of PNL's current develop-
ment effort is the destruction of hexone. Hexone, also known as
methyl isobutyl ketone MIBK and 4-methyl-2-pentanone, has the
following chemical structure: (CH3-2CHCH2COCH3. The hexone
waste was generated by an actinide separation facility operation
and it is lightly contaminated with iodine-129 and  rare-earth
elements. Some of the tanks in which the hexone is stored also
contain water and normal paraffins.
  The current approach taken by PNL uses an acid and metal ion
solution and an electrochemical cell to oxidize hexone to carbon
dioxide and hydrogen. To date nitric acid solutions of nickel (III)
and cobalt (III) have been used with good success.

INTRODUCTION
  Researchers at PNL are pursuing the concept of electrochemical
oxidation of hexone to carbon dioxide and hydrogen. Initially,
work was focused on using sulfuric or nitric acid solutions of silver
(11) and/or cobalt  (III). This combination can readily destroy
hexone and theoretically is capable of destroying tars and normal
paraffins. One objective of this work is to develop a safe method
of destroying 125 000 1 of contaminated hexone waste stored in
tanks at Hanford. This hexone waste is left over from operation
of a reduction, oxidation facility and is lightly contaminated with
iodine-129 and rare-earth elements.
  Electrochemical oxidation of this waste offers several advantages
over other methods  such as incineration or distillation. Metals
contained in the hexone or aqueous phases will be solubilized with
no chance of aerosol formation. In an electrochemical system
employing silver, iodine will react with the silver and precipitate.
The silver iodide can then be filtered and removed. The Hanford
tanks contain a total of about 19 g of iodine-129. If this iodine
is evenly distributed, its concentration is about 1.2 x 10~6 M.
Silver iodide is extremely insoluble with a solubility product of 8.5
x 10~17. A modest silver concentration of 0.01 M in the anolyte
solution will result in an iodine concentration reduction to about
8.5 x 10~15; essentially all of the iodine would be removed. Final
hexone reaction products will include hydrogen gas and carbon
dioxide. The hexone will be removed from the waste tanks in small
portions and destroyed. The amount of hexone that will be out
of the tank and in the hexone destruction system at any one time
will be very small.
  This paper discusses  the approach taken in using the electro-
chemical oxidation method, the results from using the method in
three chemical systems, the theory behind the method, and recom-
mendations and conclusions.

EXPERIMENTAL APPROACH
  An electrochemical flow cell was set up in a manner similar to
that shown in Figure 1. This configuration was used to conduct
most of the experiments. The flow cell was an Electro-Synthesis
Company Microcell with a stainless-steel cathode and a platinum
on titanium anode. A Riapore 1035 anionic membrane separates
the anode and cathode. The anode and  cathode each had an
available electrode surface area of 20 cm2. The anolytes used in
the experimental work included a metal ion, such as silver, cobalt,
manganese, or nickel, dissolved in 4 N nitric acid. Other nitric acid
concentrations were used, and 4 N sulfuric acid was also used with
cobalt. The catholytes included equal concentrations of sulfuric
and nitric acids.
             All Tubing 1/4" Teflon
             Except Hexone Addition
             1/8" Tellon and Tygon
                                                 H! To Gas
                                                  * Collector
                 Sample
                            Sample
                 All Teflon Pump Head 0-2.5 LPM
                          Figure 1
      Simplified Electrochemical Oxidation System Configuration
                                                                                                    TREATMENT    183

-------
  The anolyte and catholyte in the system each had a volume of
300 ml. The anolyte was heated and stirred. A burette (and later,
a separatory funnel) was inverted in a large beaker of water to
measure  the volume of gas evolved. Flow through the cell was
provided by two Teflon  pumps (Saturn model SP 2000 with
Minarik motor controllers). Direct current power was supplied by
a  Hewlett-Packard  model 6281A  DC  power supply with  a
maximum capacity of 6 amp. Teflon tubing was used to connect
all of the system components.  Galtek sample valves were used to
extract samples while the  system ran.
  During a typical  experimental run,  300 ml  of anolyte  and
catholyte were added to the cleaned and leak-checked system. The
circulating pumps heaters and stirrer were turned on, and the system
was brought to the desired temperature. Temperatures of 35 and
50C were used in the experimental runs. When the system was up
to temperature,  5 ml of hexone  were added and allowed to mix
with and fully dissolve into the anolyte. An initial sample was taken
before the power supply was turned on to serve as the basis for
subsequent hexone concentration determinations. The power supply
was then turned on and adjusted to provide 6 amp to the cell; the
time, voltage and amperage were noted on the data sheets. Samples
were  taken at 30-min intervals  for gas  chromatographic  (GC)
analyses. Gas evolution from the catholyte and anolyte was noted
and gas  samples were  taken for GC analysis.
   Gas chromatography was the primary  analysis tool. The most
significant source of error is a random error associated with the
injection of the sample into the gas chromatograph. The syringe
injects microliter samples, and sample sizes can vary. The maximum
size of this error is estimated to be on the order of 5%. All of the
graphs and indeed all of the hexone destruction data are based on
GC analysis. Gas chromatography was especially difficult for the
sulfuric acid samples. It was discovered that sulfuric acid attacked
the organic packing in the column, which led to spurious results.
To deal with this difficulty, the  sulfuric acid samples  were
neutralized before  injection in  the gas chromatograph.  Upon
neutralization, a precipitate formed that had to be removed by cen-
trifuging before injecting the sample into the gas chromatograph.
This neutralization step, combined with the greater time lag between
anolyte sampling and GC analysis,  resulted in more data scatter
for the sulfuric acid system.

RESULTS
   Hexone  is readily destroyed by silver and nickel  in nitric acid
and cobalt in either nitric or sulfuric acid. The use of sulfuric acid
  400-


  350


  300-

S




I 200



1 1 50
I


  1 00-



  050-
                                          * 3 M, 50'C
                                          X 3 y. 50ฐC. 2 M, HNOj
                                          • 3 M. 35ฐC
                                          1 0 75 M. 90ฐC
                                          D 0 76 M 35ฐC
                50
                        too
                                  150
                               Tim0 (mm)
—I—
 200
                                                  —I—
                                                   250
                                                             300
                           in the catholyte and silver in nitric acid in the anolyte was the first
                           choice of researchers at PNL to avoid the formation of NOx in
                           the catholyte. However, the sulfate ion migrated through the
                           anionic membrane, precipitated the silver in the anolyte and clogged
                           the cell. When nitric add was used in the catholyte with either silver
                           or cobalt dissolved in nitric acid in the anolyte, no precipitates
                           formed at low hexone destruction rates. In a preliminary test where
                           hexone was continuously pumped into the reaction vessel, a white
                           precipitate formed. This test used a combination of silver and cobalt
                           in nitric acid in the anolyte and achieved a high specific hexone
                           destruction rate. Additional continuous tests are in progress and
                           results will be reported  in later publications.
                             400
                                                               1 O.I M. 60*C. 0 5 H, MNOป
                                                               • 4 M. 60*C
                                                               • 1 M. 5O-C
                                                               0 0 I M 35'C
                                                               x 1 M. 36*C
                                                               • 4 M. 3S*C
                                                      Figure 3
                                Hexone Destruction Versus Tune: Cobalt in 4M Nitric Acid
                             350-1	
                                      T 0 5 M, 5O-C
                                      ซ 2 M. 35'C
                                      X 0 05 M. 6O*C
                                      ป 0 05 y. 36'C
                                                                                                      150       200

                                                                                                    Ttmซ (mm)
                            Figure 2
      Hexone Destruction Versus Time: Nickel in 4 M Nitric Acid
                                                      Figure 4
                             Hexone Destruction Versus Time: Cobalt in Normal Sulfuric Acid
                             Results for three chemical systems are presented graphically in Figures
                           2 through 9. The three systems are cobalt in nitric acid cobalt in sulfuric
                           acid and nickel in nitric acid. Work also was conducted with manganese
                           in nitric acid, with poor results. The GC analysis data were converted into
                           .hexone concentrations and plotted against time in Figures 2 through 4 and
 184    TREATMENT

-------
  4.00
                                              + 3 M. 50ฐC
                                              X 3 M, 50ฐC, 2 M. HNOj
                                              0 3 M. 35ฐC
                                              0 0.75 M, 50ฐC
                                              • 0.75 M. 35ฐC
  0.00
                                                                         3.50
                                                                          2.50-
                                                                          2.00-
                                                                          1.50-
                                                                          1.00-
   0.50-
                                                                          0.00
                             Energy Input (whr)
                                                                                   Y 0.5 M. 50ฐC
                                                                                   0 2 M. 35ฐC
                                                                                   X 0.05 M, 50ฐC
                                                                                   + 0.05 M. 35ฐC
                                                                                     20       40       60      80      100      120      140
                             Figure 5
  Hexone Destroyed Versus Energy Input: Nickel in 4 M Nitric Acid
                             Figure 7
Hexone Destroyed Versus Energy Input: Cobalt in Normal Sulfuric Acid
   4.00
                                        0.1 M. 50ฐC, 0.5 M, HNOa
                                      o 4 M, 50ฐC
                                      ซ 1 M, 50ฐC
                                      D 0.1 M. 35ฐC
                                      X 1 M. 35ฐC
                                        4 M. 35-C
   0.00
                                                                150
                               Energy Input (who
                                                                                     3 M. 50ฐC
                                                                                   X 3 M, 50ฐC, 2M,HN03
                                                                                   D 3 M,35ฐC
                                                                                   I 0.75 M. 50ฐC
                                                                                   0 0.75 M, 35ฐC
                             Figure 6
   Hexone Destroyed Versus Energy Input: Cobalt in 4 M Nitric Acid
                            Figure 8
   Concentration Versus Destruction Rate: Nickel in 4 M Nitric Acid
against energy input in Figures 5 through 7. The differential rate of hexone
destruction was determined by dividing the difference in the hexone con-
centrations by the sample interval and multiplying by the volume of anolyte
solution. Figures 8  and 9 show this differential destruction rate plotted
against the average hexone concentration over the time interval.
  Figures 2 through 4 show cumulative hexone destruction versus
time for nickel in nitric acid, cobalt  in nitric acid and cobalt in
sulfuric acid. The nitric acid results shown in Figures 2 and 3 are
similar, and the data sets in  these graphs correlate well with a
second-order fit. The nitric acid results are considerably better than
the sulfuric acid results shown in Figure 4. The data for the cobalt
in sulfuric acid system in Figure 4 did not correlate well with a
first, or second-order fit. A first-order fit is shown as a reference.
  Figures 5 through 7 show the same cumulative hexone destruc-
tion versus cumulative energy input. The nitric acid graphs shown
in Figures 5 and 6 are similar, and data sets  correlate reasonably
well with a second-order fit. The data for the nickel in nitric acid
(Fig. 5) show a much wider range than the data for cobalt in nitric
acid (Fig. 6). Both of the nitric acid anolytes show a substantially
higher electrical efficiency than the sulfuric acid results shown in
Figure 7. Again the sulfuric acid results did not correlate well with
a second-order fit, and a first-order fit is shown for reference.
  Figures 8 and 9 show the differential destruction rate versus the
average hexone concentration over the time interval during which
the differential destruction rate was determined. The data have a
wider scatter because of their differential nature. However, the bulk
of the data from the nitric acid runs  shown in Figures 8 and 9
indicate a positive correlation between hexone, the rate of destruc-
tion and the hexone concentration. The data from the sulfuric acid
test contained too much scatter to generate a meaningful plot.

THEORY
   A preliminary theory on how the electrochemical system attacks
and oxidizes hexone in the nitric acid is as follows: at the anode
surface, the metal ion is oxidized to a higher valence state and the
nitrate ion is  oxidized to nitrate radical as shown in Equations 1
and 2.
                                                                                                                TREATMENT     185

-------
NO3-     NO3<
                                                          (1)
                                                          (2)
    In solution, the species produced at the anode surface further
 react as described in Equations 3 and 4. Equation 5 is the net result
 of the sequence to this point.
    NO
    H2O
H2O     HNOj  + OHฐ

 NO,-    NO3ฐ + M + +
H * + OH ฐ (net reaction)
                                                      (3)
                                                      (4)
                                                      (5)
   The hydroxyl radical is highly effective in abstracting the tertiary
 hydrogen from the hexone as depicted in Equation 6.
   500


   450-


   400-


|  3 SO-

9
•  300-
c
|  JSO-

i
;  200-
S
   I SO-


   I 00-


   050-
  OOO
       f 0 1 M, SO"C. 0 &
       ป 4 M, SO*C
       • I M. SO*C
       O 0 1 M. 35ซC
       ซ 1 M 3S*C
       * 4 M 35-C
                              6        B
                             * Concentration (g -'
                                             10
                            Figure 9
   Concentration Versus Destruction Rate: Cobalt in 4 M Nitric Acid
   OHฐ +  HEX     H2O + HEX"                        <6>

The hexone radical then cleaves and reacts with the solution to form
two molecules of acetone as the major by-product. Acetone is
clearly evident in the GC traces and is readily degraded to carbon
dioxide and acid by nitric acid.

CONCLUSIONS AND RECOMMENDATIONS
  The volume of 125,000 I of hexone is equivalent to 100,125 kg
of hexone. Using the experimental results, a preliminary estimate
of the energy, equipment and time requirements to destroy the
hexone can be made. Based on typical results, 1.67 million kWh
are needed. At a cost of $0.04/kWh, this results in an electrolyzer
energy cost estimate of $66,750. (Note that this cost is a direct scale-
up of laboratory results.) Typical electrochemical  cells can have
a maximum current density of about 0.4 amp/cm2. Assuming a
4-V operating potential and a 6-month operating campaign, the
required cell anode size would be about 24 m2. Calculations show
that three full-size commercial cells, each with 8 m2 of anode
surface  area, could  do the job in about 6 months. The full-size
cells cost $80,000 to S100 000 each. Ancillary system  piping pumps.
controls and power supply would cost about $100,000. The net
result is total equipment and  energy costs of approximately
$406,750. The envisioned system would be skid mounted so that
it could be used at other Han ford sites and possibly at other U.S.
Department of Energy sites.
  The primary recommendation is that electrochemical oxidation
be considered as a method of disposing hexone and other wastes
at Han ford. Process development work should be conducted to
verify the required size of the overall process. A single cell with
0.4 m2  of anode surface area should be tested  with a variety of
electrolytes and actual hexone waste. This single cell costs about
$20,000 and is designed to be expanded into the full-size unit. Scale-
up is  exactly linear; thus, most of the equipment used to verify
process performance can be used in the full, scale system.

ACKNOWLEDGEMENT
  This  work was done at the  Pacific  Northwest Laboratory,
operated for the U.S. Department of Energy by Battelle Memorial
Institute under Contract  DE-AC06-76RLO 1830.
186    TREATMENT  ,

-------
           Advanced  Chemical  Fixation  of  Organic  Content Wastes
      In  Conjunction with  Japanese  Ground  Engineering  Equipment
                                                  Jeffrey P. Newton
                                         International Waste Technologies
                                                   Wichita, Kansas
ABSTRACT
  The first segment of this paper discusses an explanation of an
advanced chemical fixation technology for the treatment of in-
organic and organic toxic wastes. A major focus is on the treat-
ment of high content organic wastes. There is a discussion on the
treatment of PCBs in soil.
  The second segment of the paper presents the JST mixing  drill
for in situ treatment. This Japanese injection and mixing system
will be used in a SITE project in Florida for the treatment of PCBs
in soil. Other Japanese treatment and containment systems are
reviewed.

INTRODUCTION
  This paper discusses a chemical fixation technology that  uses
sophisticated silicate-based chemicals to generate chemical bonding
mechanisms that have the ability to tie-up or bond wide ranges
of mixtures of organic and inorganic toxic wastes that are in the
form of liquids, sludges, soils and sediments. The objective of this
chemical fixation approach is to effectively and simply treat wastes
at lower costs to prevent unacceptable leaching of the toxics.
  Initially we discuss how this inorganic polymer system functions
as a medium in which various toxic organic compounds are tied-
up, this speculation is based on various TCLP, modified EP toxicity
leach and solvent extraction tests. We have also based our analy-
sis on current state-of-the-art interpretations of cement reaction
mechanics from British and Japanese sources. These initial theories
are examined in light of infrared scanning experiments of untreated
and treated samples of nitrobenzene, triethonolamine, phenol, and
4- chloroaniline by the HWT-22* chemical fixation compound.
Thermogravimetric analysis of untreated and treated triethanola-
mine and phenol were performed to support the conclusions of
the infrared spectra data.
  There are a number of degrees of novelty associated with  this
process:

• A conceptual novelty in that unique inorganic polymer or sili
  cate colloid-based environments can be used as a durable medium
  to achieve  a wide range of bonding types with a wide range
  organic substances.
• The intermediate and end-state morphology, microstructure  and
  chemistry of the inorganic substrate is formulated to be rela-
  tively durable and irreversible and promote absorption of the
  organic molecules and colloids into a complex silicate maze
  whereby a sufficient bond of some type will occur. This is much
  different from the  adsorption process of earlier  generation
* Patent Pending
  solidification technologies.
• The use of admixtures is unique in a two dimensional chemical
  sense: they affect the internal cement hydration reaction to cause
  a more effective dispersion of treatment chemicals throughout
  the waste medium
• They externally promote certain beneficial surfactant functions
  in the waste medium that promote a microscopic homogeneity
  of mixing.
• There is the use of intercalation  or linking compounds that
  interact with the organic toxic compounds by a sorptive process
  in the initial stage of the treatment reaction. The nature of this
  bonding ranges from Van der Waall's forces to delta,  pi and
  covalent bonds. These intercalation compounds can bond into
  the foundation silicate substrate.

OVERVIEW OF HWT-22 CHEMISTRY
  A fast sorption interaction occurs between the intercalation com-
pounds and organic compounds in the treatment medium. Later,
a bonding takes place between the  linking compound and the
silicate-based macromolecule matrix. This intercalation compound
is a sodium magnesium fluorolithosilicate and/or sodium bentonite
that has been reacted with a quaternary ammonium compound to
make it reactive to a broad base of organic compounds. The nature
of bonding between the intercalation compound and the toxic
organics ranges from weak Van der Waall's forces to strong delta,
pi and covalent bonds.
  These processed smectites, mentioned  above, also  have the
highest cation and anion exchange capability. This ion exchange
capacity is helpful in a number of chemical interactions with
organics as well as inorganics. The crystal morphology, number
and type of elements embedded in the octahedral and tetrahedron
units is significant in its ability to chemically interact with a variety
of compounds and elements. The smectites have the ability to
expand along the c-axis to sorb a large number of toxic organic
compounds. All of our recent experimental evidence continues to
support the above thinking on the Phase I reaction mechanism
(Fig. 1)
Silicate Medium
  The generation of the inorganic polymer macromolecule matrix
is also a relative irreversible colloid synthesis. However, this is  a
slower reaction going from sol, to gel, to an irregular, crystalline,
three-dimensional, silicate-based medium. In  the formulation of
this unique cementitious material, a number of chemical issues are
considered:

• The morphology or form of the individual particles throughout
                                                                                                  TREATMENT    187

-------
         O Oiygenj   ฉ Hyd,,ปrl--    • Alununum. .too. m.jrwiium

         O tnd • Silicon. oco*SMtii iiiv ilouininin
                                                                   HWT-22 mix design. No one component has a majority effect on
                                                                   the ability of the mix design to stop the leaching of toxic organic
                                                                   compounds.
                                                                                              Tibfe 1
                                                                                   Uปch and Extraction Tetl Rrculu
                                                                         ซulซ* r*r MvT.M C*MlMl
                                                                                                      UlVปMR|ซ
                                                                  Oil,
                                                                   llซwl

                                                                  Oily
                                                                                                       ซ•••ซปป       e.j
                                                                                                       (.11 > .....i  o.n
                                                                                                      C* %••!•!ซ,'    ••
                                                                                                ll*v (M VMirua rilw 'w V* 1
                            Figure 1
                   A Particular Smectite Structure
   the reaction sequence
 • The microstructure of the body of particles and the way in which
   they are linked
 • The development  and  density of the sulpho-ferri-hydrates
 • The durability of the end-state structure to long term exposure
   to  depreciating environmental conditions

   Certain  unique  admixtures are used to cause  a  more even
 microscopic distribution of polymer-producing materials through-
 out the toxic waste that  is being treated. These admixtures also
 have  surfactant capabilities  in terms of wetting, dispersing and
 emulsifier  functions. These admixtures are designed to function
 in  relatively impure environments.  The  type  and amount of
 admixture used takes into account the type of background extender
 (fly ash, kiln  dust, soil,  etc.) and the nature of the waste. The
 admixtures used in the HWT compounds promote better develop-
 ment of the Interpenetrating Polymer Network (IPN) bonding as
 a result of this more uniform microscopic distribution of HWT
 compounds. IPN bonding is an entanglement of molecules cor-
 responding to at least a weak bond. However, bonding of much
 greater strength can  also occur under the  right  circumstances.
   The initial speculation  that chemical bonding  is occurring was
 based on leach and  extraction tests of various  types of treated
 organic waste  and a  review of various sources of information on
 inorganic polymer chemistry, advanced cement science as it applies
 to new materials development, inorganic/organic bonding theory,
 etc. Subsequent infrared absorption and thermogravimetric analysis
 support the theory that significant bonding is occurring. We also
 believe bonding of a given compound in  a  HWT-22  treatment
 medium  can vary  according to the nature of the  overall waste
 environment. This conclusion is based on extraction experiments
 with PCBs and pentachlorophenol in untreated  and treated soil
 (Table 1)
   There  is  a balance among all the functional components of the
LEACH AND EXTRACTION TEST RESULTS
  A sludge sample containing 70% water with the following toxic
components was treated at 15% by weight of HWT-20. the leaching
procedure was TCLP except for cyanide. Cyanide was leached by
deionized water.
 Chemical

 Acrylonitrite
 Acrylic Acid
 Acrolein
 Acetonitrile
 Copper
 Antimony
 Organic cyanide
 Free cyanide
  Total in
Solid (ppm)

    120
     5
    59
    150
    78
    13
    120
    10
Leachaokin
Solid (mg/l)

   1.5
   0.1
   0.5
   3.9
   0.2
   0.7
   0.13
   0.10
  The following liquid waste was treated 33% by weight with
HWT-20. The total organic content was 68%. Using the TCLP
leach test procedure the following results were achieved. Sample
cure time was 3 to 4 weeks.
Chemical                    Total (ppm)


Vinyl Chloride                    1,671
Trichloroethylene (TCE)           11,200
Trichloroethylane (TCA)            3,800
Tetrachloroethylene (PCE)          5,900
N.D. =  Not Detectable
                   LeacHable
                     N.D.
                     N.D.
                     N.D.
                     N.D.
  The tested API separator bottoms were described as filtering day
contaminated  by oil and grease (24.9%) along with  the con-
taminants listed. HWT-22 treatment was at 15% to 25% by weight.
Using the TCLP procedure, most chemicals were not detectable
(ND) in the leachate with the detection limit equaling 0.1 mg/l.
188    TREATMENT

-------

Chemical
Chromium
Lead
Ethyl benzene
m-Xylene
o- & p-Xylene
Anthracene
Chrysene
Methylnaphthalene
Naphthalene
Phenanthrene
Solid
Concentration (ppm)
630
250-332
10
40
43
19
29
170-470
13-93
110-206
                                             Leachable (mg/l)

                                                0.03-0.04
                                                  0.05
                                                  ND
                                                  ND
                                                  ND
                                                  ND
                                                  ND
                                                  ND
                                                  ND
                                                  ND
   As a means of demonstrating that variations in HWT formulas
affect leach test results, the following example of treatment of
refinery wastes is given. This waste was chosen because it is a very
difficult waste  for standard chemical fixation to treat because of
its very high organic content.  The waste used in this test with
HWT-20 treatment compound was a liquid with some solids; prin-
cipally the was K049-Slop Oil Emulsion, with some K048-DAF
Float and K051 API Separator Bottoms. This mixture is a combi-
nation of refinery wastes. Our initial treatment was as follows,

•  The waste was mixed on a one to one weight basis with cement
   kiln dust which was used as  a thickener
•  The waste was mixed with HWT-20 in slurry form with 15 Ib
   by weight of HWT being used for every 100 Ib  of the original
   oil emulsion.
•  The product  was capped and cured 21 days before leach testing.
•  The volume expansion was about 1.25. Since there was no water
   in the waste  to begin with, the majority of the volume expan-
   sion was caused by the water in the slurry.
  The following are the Toxicity Characteristic Leaching Procedure
results:
                                        Extraction by varying polarity solvents of same treated sample
                                      discussed in previous example, API Slop Oil Emulsion yielded the
                                      results below:
Chemical

Chromium
Lead
Arsenic
Barium
Cadmium
Mercury
Selenium
Silver
Copper
Zinc
Nickel
Vanadium
Xylenes
4-methyphenol
Isophorone
2-nitrophenol
2,4-dinitrophenol
Pentachlorophenol
Phenanthrene
Anthracene

N.D. =  Non-detectable
Total (ppm)

     1.9
    16.3
    4.81
   22.12
    2.40
    1.25
    1.25
   57.70
   32.70
    3.10
    6.25
  173.10
  26,500
    9.1
   2,226
     816
     316
     49
     21
     28
  Leachable (ppm)

          0.28
          0.08
         0.009
          1.4
          0.04
          0.08
         0.0049
          0.07
          0.06
         0.005
         0.020
          0.2
        4,605
         N.D.
          8.5
          1.3
         N.D.
          1.3
          1.4
          1.09
  As one can see, too much xylene leached out .so we had to change
the formulation. In the second leach test, we  followed the same
sample preparation procedure except HWT-21 was used as the
treatment material rather than HWT-20. TCLP results were:
Chemical

Xylenes
All other organics
Total (ppm)

  26,500
Leachable (ppm)

48
Dimethylsulfoxide n-Butanol Iso-Octane
Chemical Polarity Index Polarity Index Polarity Index
7.2 (ppm) 4.0 (ppm) 0.1 (ppm)
4-Methylphenol
Isophorone
2,4-Dimethylphenol
2-Nitrophenol
Hexachlorobenzene
Pentachlorophenol
Anthracene
Phenanthrene
Methyl-naphthalene
10.0
2179.0
207.5
724.0
145.0 ppb*
21.3
3.27
6.5
219.0 ppb*
5.5
1213.0
116.0
361.0
97.0 ppb*
13.7
4.7
3.9
122.0 ppb*
1.9
695.3
29.0
91.4
23.9 ppb*
5.6
22.8
19.2
844.0 ppb*
                                        Not reported in previous analyses due to its low levels of concentration
  An examination of the above data reveals that the greater the
polarity of the organic compound, the stronger the bondage with
the matrix.  This  result is  shown very well  in  the  case  of
4-methylphenol where a strong solvent  like dimethyl sulfoxide
(polarity under 7.2 is required to extract the molecule from the
system, whereas other  solvents with lower polarity indices (n-
butanol 4.0; iso-octane 0.1) have very low extraction affinities for
the phenol. The same trend is revealed with the other compounds,
also, indicating there is a considerable amount of dipole-dipole con-
tribution in  the bonding phenomena.

INFRARED AND THERMOGRAVDMETRIC EXPERIMENTS
  Organic compounds of four different  classes were chosen for
this experiment. They were mixed with Portland cement in the ratio
of 1:1 by weight. To 2g of this mixture 0.4g of HWT-22 in a slurry
were added. The two parts were mixed thoroughly and  then set
aside for 48 hr. After that time, the samples were dried in a vacuum
desiccator with phosphorus pentoxide for  48 hr. The samples were
then taken out and mixed with potassium bromide and pelletized
for infrared analysis.
  The infrared spectra of the untreated samples and the treated
samples were recorded with a Perkin Elmer 1710 FTIR instrument
with scanning capabilities of 4400-400 cm-1. Thermogravimetric
analysis (TGA) was performed using a Perkin Elmer TGS-2 system
equipped with a thermal analysis data system (TAD) in air.

RESULTS
  The infrared spectra of treated and untreated samples are shown
in Figs. 2 through 5. Their absorption frequency values are shown
in Table 2. Discussion of the possible types of bonding  revealed
the infrared data generated are presented below:
                          Table 2
    Comparative Characteristic Infrared Absorption Frequencies
                       Untreeted  (cm
                                      )
                                                              1530

                                                              1359

                                                              1656
                                             Bending
                                                                    C-C1  Stretch
                           N.D
                       1372

                       1315

                        5O7

                        E4O

                        698

                       1617
Treated (cm"1)
1525

1362

1661

1478

1374

1321

 486

 610

 670

1622
                                                                                                          TREATMENT     189

-------
 Untreated
                                 C-NO  Stratohing 1S3O cm
                                                                 Untreated
                                                                                                               O-H banding
                                                                                                                     1474  cm

                                                                                                                     1372  cm
                                                                                                            -1

                                                                                                            -1
                                                                                                                     1315 cm
                                                                                                                            -1
                          Figure 2
                        Nitrobenzene
                                                                 Traatad      1
                                                                                        1  S 3 Layer. of
                                                                                                               O-H  bending
                                                                                                                     147S cซ

                                                                                                                     1374 cm
                                                                                                                     1321
                                                                                                             -1

                                                                                                             -1
                                                                                                                             -1
                                                                                                               •nd S10
Untreeted
\
  c>

 -T
  CH
                               I
                              CH
                               I
                               0
                2
                '2
                      C-O  Stretch  1359 cm


                      CH  Stretch 1658 am~
                                                                                            Figure 4
                                                                                             Phenol
                                                                  Untreated
                                                                                                               C-C1  otrotch
                                                                                                                         -1
                                                                                       507

                                                                                       640

                                                                                       698
                                                                                                          -1
                                                                                                                   1617 cm
Treated

"?
   I2
               O-  CH, - H C - N:
                V    2    2    I
                 >1           CH
                               I2

                              ซป
                               O
                                     C-O Stretch   1362
                                     CH -Stretch   1661
                                      Subatrete
                          Figure 3
                      Triethanolamine
Nitrobenzene
  There is a sharp absorption in the untreated sample at 1530
cm"1; this absorption is attributed to C-N02 stretching. The ab-
sorption frequency for the same stretch in the treated sample oc-
curs at 1525 cm"1 indicating that there is a slight elongation of
the bonding. This phenomenon could be attributed to some weak
bonding between the two oxygens in the nitro group and the treat-
ment material. Since the stretching frequency is decreased by a small
margin, we believe the bonding would fall in the category of dipole-
dipole interaction and dipole  induced dipole  interaction.
                                                                  Treated

                                                                       aubatrate
                                                                                                C-C1  atretcH
                                                                                                          -1
                                                                                                    486

                                                                                                    610

                                                                                                    670

                                                                                                NH  atretch

                                                                                                    1622 en
                                                                                               -I

                                                                                               -1
                                                                                             Figure 5
                                                                                          4-Chlorolaniline
                                                  Phenol
                                                    In the case of phenol, unfortunately, the hydroxyl stretch fre-
                                                  quencies were masked by the water in the treatment sample. So
                                                  hydroxyl bending frequencies were used to  deduce the bonding
                                                  phenomena. The  hydroxyl bending in the untreated compound
                                                  occurs at  1474, 1372,  1315 cm-'. On the other hand, in the
                                                  treated compound the same bending mode occurs at 1478, 1374,
                                                  and  1321 cm"1.
                                                    The increase in the bending frequencies may be explained in the
                                                  following manner: the benzene ring can be  sandwiched between
                                                  two  layers of the  treatment material where the electronic charge
                                                  cloud in the benzene ring can form weak bonds with the two layers
                                                  of the treatment material, thus creating a slight positive charge
                                                  inside the ring.  This charges would result in  the aspiration of the
                                                  lone pair of electrons into the ring rendering the oxygen/hydro-
                                                  gen bond weaker and thereby provide a greater degree of freedom
190    TREATMENT

-------
      I TRIETHANOLAMINE
      i	 ODHWTin
                      TREATMENT
                              VTi  31.4540 ng  RATCi  20.00 dซg/ซln
en tin ATM
DATE.  B7/04/ZI
             FILC, tsoii. re
               TIHEi  111 ID
                              TEMPERATURE 
                                                                    ซ?ซ AIR ATM.    r/if, fsosa rc
                                                                    0*TEi  87/03/20   TIHEt  12i ID
                                                                                                TEMPERATURE CO
                                                                                                                             TC
                                                                                           Figure 7
 for the hydrogen. This weakened bond accounts for the slight
 increase in the hydroxyl bending absorption frequencies.

 Triethanolamine
   There are two characteristic absorptions in the untreated com-
 pound, the C-0 stretching at 1359 cm~' and CH2 stretching at
 1658 cm"1. The same absorption frequencies have shifted to 1362
 in the treated material. The increase in the stretching frequency
 indicated a very slight shortening of the bonds and this phenome-
 non may be explained as follows: the lone pair of electrons on the
 nitrogen are used in forming a "Lewis acid-base" reaction which
 creates a slight positive charge on the nitrogen atom. The bonding
 electrons between the nitrogen and carbon atom are pulled closer
 to the nitrogen resulting in shortening of the bond length. In the
 same manner, the carbon/oxygen bond is also strengthened. The
 existence of a strengthened bond explains why both  C-0 stretch
 and CH2 stretch occur at a higher frequency than in the untreated
 sample spectrum. The bonding in this case would be classified as
 a weak coordinate covalent bond.

 4-CMoroaniline
   There are two types of bonding in the molecule; the C-C1 stretch
 occurs at 507, 640 and 698 cm-1 and the NH2 stretch  1617 cm-'.
 In the case of 4-chloroanaline, we are seeing two different pheno-
 mena at the same time. In the treated compound the C-C1 stretch
 has shifted to lower frequency, namely 486, 610 and 670 cm-',
 whereas the NH2 stretch has shifted to a higher frequency, namely
 1622 cm"1. The explanation for these phenomena can be arrived
 at by  considering two types of bonding. In the case of C-C1
 bonding, the three pairs of electrons of chlorine appear to form
 weak bonds  at both sides of the substrate. These bonds would
 restrict the C-C1 stretch resulting in shift to lower frequencies. In
 the case of NH2 stretch the lone pair of electrons on the nitrogen
 will be used in a "Lewis acid-base"  reaction resulting in  the
 enhancement of the NH2 stretching frequency. In both cases, the
 bonding would be of the coordinate covalent variety.
   Our observation is in keeping with the experimental evidence.
 Further, the thermogravimetric analysis of triethanolamine reveals
 that the compound starts leaving the substrate around 250 ฐC and
 completely  boils  off around 290 ฐC.  The  boiling point  of
 triethanolamine is 277 ฐC and the fact that it does not leave the
 system until 290 ฐC reveals that there is some kind of weak bonding.
 (Fig. 6)
Thermogravimetric Analysis of Treated Phenol
  Thermogravimetric analysis of phenol reveals that the water in
the matrix boils at 100ฐC and then around 181 ฐC (boiling point
                                                                of phenol), the phenol starts leaving the system slowly instead of
                                                                boiling off with a sharp weight loss. This phenomenon can be
                                                                explained by the hydrogen bonding  which is common among
                                                                phenols.  It appears that the phenolic hydrogen is strongly bound
                                                                to the oxygen atoms of silica and alumina, thus preventing the
                                                                phenol from leaving the matrix at its normal boiling point. Further,
                                                                the final traces of phenol do not leave the matrix until 665 ฐC (Fig.
                                                                7). This result would also  indicate that the matrix has excellent
                                                                bonding properties and would not let the pollutants escape under
                                                                ambient conditions. This sample had a 1-month cure time before
                                                                this analysis was run.
                                                                CONCLUSION
                                                                  Both the infrared spectra and TGA studies indicate there are
                                                                several possible types of bonding occurring between the substrate
                                                                and the  organic compounds. The nature of this bonding ranges
                                                                from "Lewis acid-base" reaction to Van der Waall's forces. This
                                                                phenomenon is indicated very well in the case of triethanolamine
                                                                where the shift is by three units indicating weaker bonding, whereas
                                                                in the case of chloroaniline its  shifts are rather dramatic (up to
                                                                30 units).
                                                                  The nature of bonding in the case of chloroanaline tends to move
                                                                toward true chemical bonding,  whereas in all the other cases we
                                                                have discussed, it appears to be either dipole-dipole interaction or
                                                                other weak bonding such as a hydrogen bonding or a delta bonding.
                                                                Although in a majority of the cases the shifts are all small, it indi-
                                                                cates that  there is a degree of bonding (however weak it is) that
                                                                is binding the organic compounds to the substrate. It is possible
                                                                by varying the functional groups, the pH and a few other physical
                                                                parameters of the substrate that one can enhance its capability to
                                                                hold firmly a greater variety of compounds with a higher loading
                                                                density.
                                                                EQUIPMENT FOR IN-PLACE TREATMENT
                                                                AND CONTAINMENT
                                                                  The Japanese have developed a wide range of specialty equip-
                                                                ment that was used in soil and sediment improvement activities
                                                                for civil engineering purposes. This equipment can be effectively
                                                                adapted for in situ toxic waste treatment and contain.ment. One
                                                                important consideration is that this equipment has been proven
                                                                over a wide range of ground conditions for a long period of time.
                                                                Currently the plan is to use the Sanwa Kizai 1ST Mixing Drill hi
                                                                                                        TREATMENT     191

-------
 the treatment of PCBs in soil at a former transformer repair facility
 in Florida.

 JST Mixing Drill
   The JST mixing drill is a single rotary shaft injection and mixing
 system for cement slurry and or sodium silicate liquid. It is capable
 of creating a column of 2.8 ft in diameter down to 100 to 120 ft
 deep.  Two important features of this system are that it will give
 very straight or smooth sides to created columns and it can grind
 through soft porous rock. The JST system gives a high degree of
homogeneity of soil and HWT slurry which will insure a success-
ful treatment.

Tone PCW System
  This system  will create a cement-clay wall 1.25 ft to 4 ft thick
down to 400 to 450 ft deep and is only off by 1/1000 at the bottom.
It is a non-shaft system with multiple drill bits in which the drill
motors and controls are lowered into trench by cable. Drill cuttings
are excavated to the surface and the cut off wall is emplaced as
the drill is removed.
192    TREATMENT

-------
             Bioremediation  of  Contamination  by Heavy  Organics
                               At  A  Wood  Preserving Plant Site
                                                Ronald J. Linkenheil
                                          Remediation Technologies, Inc.
                                               Fort Collins, Colorado
                                                 Thomas J. Patnode
                                               Glacier  Park Company
                                                 Seattle,  Washington
ABSTRACT
  On-site treatment was chosen as the closure alternative for a creo-
sote impoundment at a Superfund site in Minnesota. This alter-
native was identified in the FS as the most cost-effective source
control measure for the site. The effectiveness of using land treat-
ment technology to detoxify contaminated soils at the site was
demonstrated in pilot-scale studies. Results of these studies were
used to develop design criteria for a full-scale treatment facility.
  A lined 3 acre treatment facility was constructed in 1985 to treat
10,000 yd3 of contaminated  soils and sludges from the creosote
impoundment. The facility has been successfully operated by ReTec
since 1986, achieving greater than 90% reduction of polynuclear
aromatic hydrocarbons (PNAs) during the first year of operation.
This paper summarizes results from the first year of treatment and
demonstrates the effectiveness of the full-scale system. Aspects of
construction and start-up of the full-scale facility are also reviewed.

INTRODUCTION
  Wastewaters from a creosote wood preserving operation have
been sent to a shallow, unlined surface impoundment for disposal
since the 1930s. The discharge of wastewater to the disposal pond
generated a sludge which is a listed hazardous waste under RCRA.
Due to groundwater contamination of the shallow aquifer at the
site by polynuclear aromatic hydrocarbons (PNAs),  the State of
Minnesota nominated the site for listing on the NPL in 1982. Since
1982, numerous remedial investigation activities have been under-
taken to determine the nature and extent of contamination at the
site. Based on the results of  these studies and extensive negotia-
tions, the Minnesota Pollution Control Agency (MPCA), the U.S.
EPA and the owner of the facility signed a Consent Order in March
1985 specifying  actions to be taken at the site.
  In general terms, the remedial actions selected by the MPCA
and U.S. EPA involve a combination of off-site control measures
and source control measures. The off-site controls involve a series
of gradient control wells to capture contaminated groundwater.
The source control measures include on-site biological treatment
of the sludges and contaminated soils and capping of residual con-
taminants located at depths greater than 5 ft. Costs for on-site treat-
ment and capping were estimated to be $59/ton.

PILOT-SCALE STUDIES
  Before the on-site treatment alternative was implemented, bench-
scale and pilot-scale studies were conducted to define operating
and design parameters  for  the full-scale facility. Several per-
formance, operating and design parameters were evaluated in the
land treatment studies. These included:
  Soil characteristics
  Climate
  Treatment supplements
  Reduction of gross organics and PAH compounds
  Toxicity reduction
  Effect of initial loading rate
  Effect of reapplication

  Three different loading rates were evaluated in the test plot
studies: 2%, 5% and 10% BE (benzene extractable) hydrocarbons.
The soils used in the pilot study consisted of a fine sand which
was collected from the upper 2 ft of the impoundment. The soil
was contaminated with creosote constituents consisting primarily
of PNA compounds. Total PNAs in the soil ranged from 1000 to
10,000 ppm, and BE hydrocarbons in the contaminated soil ranged
from approximately 2 to 10% by weight.
  Because the natural soils are fine sands and extremely perme-
able, it was decided that the full-scale system would include a liner
and leachate collection system to prevent possible leachate break-
through. To simulate the proposed full-scale conditions, the pilot
studies  consisted  of five lined, 50-ft2 test plots with leachate
collection. The studies were designed  to maintain soil conditions
which promote the degradation of hydrocarbons. These conditions
included:

• Maintain a pH of 6.0 to 7.0 in the soil treatment zone
• Maintain soil carbon to nitrogen ratios between 50:1 and 25:1
• Maintain soil moisture near field capacity

  Hydrocarbon losses in the test plots were measured using benzene
as the extraction solvent. The analysis of BE hydrocarbons provides
a general parameter which is well suited to wastes containing high

                        Table 1
           Comparison of Pilot-Scale Kinetic Data
               At Two Initial Loading Rates
                 First Order Rate Constant (day* ^)

                	5% Plol	10*. Plot
                        HaH-IHa Mays)

                     5% Ptol	10% Pint
   Benzene
     Extractable

   2-Ring PAH

   3-Ring PAH

   4-Ring PAH

   TOTAL PNAs
0.003

0.023

0.016

0.004

0.009
0.003

0.023

0.016

0.001

0.008
231

 30

 43

173

 77
231

 30

 43

693

 87
                                                                                                    TREATMENT    193

-------
molecular weight aromatics such as creosote wastes. Reductions
of BE hydrocarbons were fairly similar between all the field plots.
Average removals for all field plots over 4 months were approxi-
mately 40% with a corresponding first order kinetic constant (k)
of 0.004/day.
  The reduction of PNA constituents was monitored by measuring
decreases in 16 PNA compounds. The following compounds were
monitored in the test plots:
2 Rings
Naphthalene
Acenaphthylene
Acenaphthene
3 Rings
Fluorene
Phenanthrene
Anthracene
                                   4, 5 and 6 Rings
                                   Fluoranthene
                                   Pyrene
                                   Benzo(a)anthracene
                                   Chrysene
                                   Benzo0)fluoninlhene
                                   Benzoflc )fluoranthene
                                   Benzo(a)pyrene
                                   Dibenzo(a,h)tnthrecene
                                   Benzo(g,h,i)perylene
                                   Ideno(l,2,3lc^l)pyrene

  Greater than 62% removals of PAHs were achieved in all the test
plots and laboratory reactors over a 4-month period. PAH removals
for each  ring class are shown below:

• 2  ring PAH: 80-90%
• 3  ring PAH: 82-93%
• 4+ ring PAH: 21-60%
• Total PAH: 62-80%

  Table 1 summarizes first order rate constants and half-life data
for BE hydrocarbons and PNA compounds for the 5 and 10% BE
hydrocarbon test plots. With the exception of the 4 and 5 ring
PNAs, the table shows that the kinetic values are approximately
equal for the 5 and 10% loading rates. In the case of the 4 and
5-ring compounds, the 5% loading rate resulted in higher kinetic
rates for  these compounds as compared to the 10% loading rate.
This difference may have been due to the availability of more 2-ring
and 3-ring compounds to soil bacteria at the 10% loading rate.
These compounds may be preferentially degraded by soil bacteria.
OPERATING AND DESIGN CRITERIA
  The pilot-scale studies were successful in developing operating
and  design criteria for a full-scale  system. These criteria  are
summarized below:

• Treatment period can be extended  through October
• Soil moisture should be maintained near field capacity
• Soil pH should be maintained between 6.0 and  7.0
• Soil carbon:nitrogen ratios should be maintained between 25:1
  and 50:1
• Fertilizer applications should be completed in small frequent
  doses
• Initial benzene extractable hydrocarbon contents of 5 to 10%  are
  feasible
• Waste reapplication should occur after initial soil concentrations
  have been effectively degraded
• Waste reapplication rates of 2 to 3 Ib of benzene extractable per
  ft3  of soil  can  be  effectively degraded during  the  3 months
  degradation occurs

  The studies suggest that all the loading rates tested are feasible.
First order rate constants were fairly  similar between all the test
plots although the intermediate loading rate (5% benzene extract-
able hydrocarbons) may demonstrate a slightly higher removal of
high molecular weight  PAH compounds. The higher loading rates,
however, showed the greatest mass removals. Selecting an initial
loading  rate should balance additional  land area requirements
against time  requirements to complete the treatment process.
Moderate loading rates (5%) will result in a  faster detoxification,
whereas higher loading rates will decrease land area requirements.

CONSTRUCTION AND START-UP OF FULLSCALE SYSTEM
  Construction of the full-scale system involved preparation of a
treatment area within the confines of the existing RCRA impound-
ment (Figure 1). The treatment area was constructed on top of the
impoundment to avoid permitting a  new RCRA  facility. If the
facility was located outside the impoundment, then a Pan B permit
                                                            Figure 1
                                               Site Plan for On-Site Treatment System
194    TREATMENT

-------
would have to be obtained before constructing the treatment facility.
By locating the treatment area within the confines of the impound-
ment, the treatment system was considered part of closure of the
impoundment. This enabled us to accomplish the cleanup quickly
and avoid the delays associated with permitting a new RCRA unit.
  The principal construction  activities at the site involved:

• Preparation of a lined waste pile for temporary storage of the
  sludge and contaminated soil
• Removal of all standing water in the impoundment
• Excavation and segregation of the sludges for subsequent free
  oil  recovery
• Excavation of approximately 3-5 ft of "visibly" contaminated
  soil from the impoundment  and subsequent storage in the lined
  waste pile
• Stabilization of the bottom of the impoundment as a base for
  the treatment  area
• Construction of the treatment area including installation of a
  100 ml HOPE liner, a leachate collection system and 4 ft of clean
  backfill
• Installation of a sump for collection of the stormwater and
  leachate
• Installation of a center pivot irrigation system
  As  previously discussed, a lined treatment area was constructed
because the natural soils at the site are highly permeable. A cap
also was needed for the residual contaminants left in place below
the liner. Therefore, the treatment area liner serves two functions
at the site: (1) to provide a barrier to leachate from the treatment
area and (2) to provide a cap over the residual contaminants that
were left in place.
  The treatment area was constructed on top of the existing
wastewater  disposal pond  after removal of all contaminated
materials. The surface area for treatment is approximately 125,000
ft.2. Containment berms with 3 to 1 slopes enclose the treatment
area and prevent surface runoff from leaving the site.
                                                      The treatment area is lined with a 100 mil HDPE membrane
                                                    (Figure 2). The base of the liner slopes 0.5% to the south and west.
                                                    A sump with a 50,000 gal capacity is located in the southwest corner
                                                    of the treatment area. A layer of silty sand ballast 18 in. thick was
                                                    placed on top of the treatment area liner. A 6-in. gravel layer was
                                                    placed on top of the ballast to serve as a leachate collection system
                                                    and as a  marking layer for land treatment operations.
                                                      The leachate collection system includes 2-ft wide leachate col-
                                                    lection drains at 100-ft centers (Figure 2). The drains are filled with
                                                    gravel and perforated pipe to  carry leachate from the collection
                                                    system to the sump. The drains were wrapped in filter fabric to
                                                    prevent clogging. A 2-ft layer of uncontaminated sand was placed
                                                    above the leachate collection system. This layer of sand serves as
                                                    an initial  mixing layer for the contaminated soils and is the treat-
                                                    ment zone for the full-scale system.
                                                      Water in the leachate collection sump is discharged by gravity
                                                    flow to a  manhole and is automatically  pumped via a lift station
                                                    to a 117,000-gal storage tank. Water in the storage tank is recycled
                                                    back to the treatment area via a spray irrigation system. Water in
                                                    excess of  irrigation requirements is discharged to the municipal
                                                    wastewater treatment plant.
                                                      Construction of the waste pile and treatment area was completed
                                                    in October 1985. In late April 1986, a center pivot irrigation system
                                                    was installed and 120 tons of manure were spread in the treatment
                                                    area. Manure loading rates were based on achieving a carbon:nitro-
                                                    gen ratio  of 50:1.  In addition to nitrogen, the manure provides
                                                    organic matter which enhances absorption of the hazardous waste
                                                    constituents.
                                                      In May 1986, a 3-in. lift of contaminated soil was applied to the
                                                    treatment area.  The target loading rate  for startup was a BE
                                                    hydrocarbon concentration of 5%. The soil was mixed (rototilled)
                                                    with 3 in.  of native soil to achieve a treatment depth of 6 in. This
                                                    application involved approximately 1200 yd3 of sludge and con-
                                                    taminated soil. The data for the startup of the full-scale facility
                                                    are summarized in Table 2.
                                                                LIMCft B6DOINC,. MINIMtli
                                                                OF SILTV *A*Jo riAceo IN •• LIFT*
                                                                ANO CarriPACTeo Tt> ISf. AVTMI Dft
                                                   SECTION C-C
                            --TW Of flttAL Aซ"-lCATUปr4
                         V
                                       \     C"""n
                                        v   '-I  	
                         Sk.       8>.X      M:

                                I   C &\   *ILTV ftAND  	L_
                        ปT      . .^^ I . . - - -   LINUE BALL>ซT
                            ^
                                      :^_
                                                                                               \fcteki
                                                                                                                 l Of H*Y TMNCH. ซซAVATฃ
                                                                                                              I' DซeP IKT6 fttCkl CCMT.COMrACT
                                                                                                              ftACtFiu. IH  i" LIFTS to ซ%
                                                                                                              **TM OH*.
TREATMENT  AKEA
                                                 DBTAlU
LIMEP  KEY  TgeMCH  DETAIL
                                                              Figure 2
                                            Cross-Section and Liner Details for Treatment Area
                                                                                                            TREATMENT     195

-------
                         Table 2
             Summary of Start-Up Data (5/23/86)
  Parameter
Average
  Benzene Extractables, %
  TOC. ppm
  TKN. ppm
  Ammonia, ppm
  Total Phosphorus, ppm
  Total Potassium, ppm
  PH
  53000
  29710
  1367
   2.37
   522
   502
   7.66
PERFORMANCE OF THE FULL-SCALE FACILITY
  Benzene extractable hydrocarbons and 16 polynuclear aromatic
compounds are being monitored to evaluate the performance of
the facility. Figure 3 shows the BE hydrocarbon concentrations
measured in the Zone of Incorporation (ZOI) during the first year
of treatment. BE hydrocarbon concentrations decreased approxi-
mately 60% over the first year of operation. Most of the decrease
occurred during the first 120 days (May through September). Little
decrease in BE hydrocarbon concentrations was observed during
the Fall and Winter months.
Polvnuclear Aromatic Hydrocarbons (PAH), ppm:
Naphthalene 1148
Acenaphthylene 21
Acenaphthene 1082
Total 2-Ring PAH 2251
Fluorene 1885
Phenanthrene 4190
Anthracene 3483
Total 3-Ring PAH 9558
E
o.
c 3000
JO
I
c 2000
ง

" 1000

0





• B

JH





I
\
\











0
Fluoranthene 1 575
Pyrene 958
Benzo(a)anthracene
and Chrysene 837
Total 4-Ring PAH 3370
Benzofluoranthenes 368
Benzopyrenes 294
lndeno(1 23cd)pyrene 1 1 1
Dibenzo(ah)anthracene 100
Benzo(ghi)perylene 106
Total 5- Ring PAH'S 979
Total PAH's 16159










n

90

Kl nUHktylllUIBIIB
• C|i tArano
riuorene
ฃ3 Phenanthrene
D Anthracene






rP

180 360








nTL













Days after First Waste Application



Figure 4



2-Ring and 3-Ring PNA Degradation vs Time







Figures 4 and 5 show PNA concentrations measured in the treatment
facility during the first year of treatment. Figure 4 summarizes data for
2-ring and 3-ring PNAs.
Figure 5 summarizes data for the 4-ring and 5-ring
compounds. Greater than 95% reductions
for the 2 and 3 ring PNAs.
in concentration were obtained
Greater than 70* of the 4-ring and 5-ring PNA
compounds were degraded during the first year of operation.


The treatment area is irrigated almost daily due to dry weather
during the summer months. Irrigation needs are determined from
soil tensiometer readings, soil moisture analyses and precipitation Q.
and evaporation records. Typical irrigation rates range from 0.25 ^
to 0.38 in. per application. This application rate keeps the soils in g
the cultivation zone moist without saturating soils in the lower treat- •> -| rjOO -
ment zone. Maintaining soil moisture near field capacity was a key 5
operating parameter in the pilot-scale studies.


5-
ฃ
0) 4-
2
a 3-
c
Qi
8 2-
0)
Q.
1 -

Q-
^ Summer ^
Months



Winter
"*~ Months "*"





8
c
5

0 -















• Fluoranthene


1 • Pyrene
• • Benzanthracene
• and Chrysene
• %2 Benzofluoranthene
•L
m
•i
m
m
_n

0
i


L
\
&






r
L



mm
•
fc
i L
[Jib

90
Q Benzopyrene



mJ% _l
-, IL i
-} im^ 1





i
k
m







LL.

180 360
Days after First Waste Application




4-Ring and





Figure 5






5-Ring PNA Degradation vs Time





With the exception of anthracene, all the 2-ring and 3-ring com-
pounds were degraded below or near detection limits after 90 days
            0     60    120    180    240    300
                  Days after First Waste Application
                         Figure 3
             BE Hydrocarbons Degradation vs Time
      360
 of treatment. Greater than 92% of the anthracene present in the
 waste was degraded during the first 90 days of treatment. Similarly,
 most of the 4-ring and 5-ring removals occurred during the first
 90 days of treatment. This result was expected because the warmest
 weather occurred  during the this period.
196    TREATMENT

-------
  Table 3 shows average PNA removals measured in the pilot-scale
studies and compares them with the full-scale removal efficiencies.
Full-scale removal efficiencies were higher than test plot removal
efficiencies for every  PNA  ring  class and BE hydrocarbons.
However, it must be noted that the full-scale facility operated for
360 days compared to only 126  days for the test plot units. Table
3 also presents average half-life data for both the test plots and
the full-scale unit. Full-scale half-lives were consistently in the low
end of the range of half-lives reported for the test plot units.

                             Table 3
           Comparison of Full-Scale and lest Plot Removals
Parameter
                    AVE. PERCENT REMOVAL
                    Ful Scale1   Test Plots2
     AVE. HALF-LIFE (DAYS)
    Fun Scale    Test Plots
2-Ring PAHs
3-Ring PAHs
4-and5-RingPAHs
Total PAHs
BE Hydrocarbons
95
95
72
90
60
199 4m el M*m
93-95
83-65
32-60
65-76
35-56
*
<45
45
115
65
150
29-33
46-49
95-226
61-83
106-202
z RmMKBdmy atuWid rfMr 18 d*> tf MMmrt.

       10000
        8000-
        6000-
        4000-
        2000
    2-Ring PAH
    3-Ring PAH
•  4-RingPAH
0  5-RingPAH
                   0           90         180         360

                       Days after First Waste Application

                             Figure 6
                  PNA Degradation by Ring Class
                              In summary, the rate and amount of PNA degradation is propor-
                            tional to the number of rings contained by the PNA compounds
                            (Figure 6). The 2-ring and 3-ring PNAs degraded most rapidly. The
                            4-ring and 5-ring PNAs  degraded at slower rates; however, these
                            compounds are strongly adsorbed to soils and are immobilized in
                            the treatment zone of the facility. Table 4 summarizes water quality
                            data  for  the  leachate collection system of the  facility.  Only
                            acenaphthene and fluoranthene were detected in the drain tile water
                            samples. Concentrations for these two  compounds  were near
                            analytical detection limits.

                                                         Table 4
                                                 Drain Tile Water Quality
                                                                                                           Concentration, ppb
                                                                       Compound
                                                       June 1986    August 1986   October 1986
 Naphthalene                    <1           <1           <1
 1-Methylnaphthalene              <1           <1           <1
 2-Methylnaphthalene              <1           <1           <1
 Acenaphthylene                 <1           <1           <1
 Acenaphthene                   <1           3.7          2.7
 Fluorene                       <1           <1           <1
 Phenanthrene                   <1           <1           <1
 Anthracene                     <1           <1           <1
 Fluoranthene                    <1           2.1           1.4
 Pyrene                        <1           <1           <1
 Benzo(a)anthracene              <1           <1           <1
 Chrysene                      <1           <1           <1
 Benzolluoranthenes              <5           <1           <1
 Benzopyrenes                   <5           <1           <1
 lndeno(123cd)pyrene              <5           <1           <1
 Dibenzo(ah)anthracene            <5           <1           <1
 Benzo(ghi)perylene              <5           <1           <1

CONCLUSION
  The data developed during this project show that on-site treat-
ment of creosote contaminated soils is feasible. Based on the data
developed in pilot-scale studies, a conservative design for a full-
scale system was developed and constructed. The full-scale unit has
matched or surpassed the performance of the pilot-scale unit in
degrading creosote organics. The advantages of on, site treatment
are that it reduces the source of contaminants at the site in a very
cost-effective manner, it  satisfies the developing philosophical
approach that the U.S. EPA has to on, site remedies and it reduces
the liability of the owner/operator that can result from off-site
disposal.
                                                                                                                TREATMENT     197

-------
                             Soil  Stabilization  Treatability  Study
                                     At  the  Western  Processing
                                                Superfund  Site

                                                    John J. Barich
                                     U.S.  Environmental Protection Agency
                                                       Region 10
                                                    Joseph Greene
                                     U.S.  Environmental Protection Agency
                                       Environmental Research  Laboratory
                                                       Rick Bond
                                                 ICF  Northwest,  Inc.
ABSTRACT
  A soil stabilization  treatability study was completed  at the
Western Processing Superfund site. Three chemical stabilization
methods and one vitrification method  were  used  to stabilize
1400 Ib.  of heavily contaminated soil. The untreated soil and
stabilized products were subjected to a series of physical, biological
and chemical performance tests. The performance test results were
evaluated for environmental effects through the use of a three
dimensional  groundwater model. The relative effectiveness of
stabilization  versus other remedial measures (no action, excava-
tion and pump and treat) was displayed through the use of a plume
animation movie.
  Chemical stabilization reduced leachate strengths to levels which
allow water quality criteria to be met directly. Leachate strengths,
as measured by the U.S. EPA's Toxicity Characteristic Leaching
Procedure (TCLP), were reduced by  "four 9's" compared to
leachates generated from untreated soil. As measured by terrestial
bioassay techniques, chemical stabilization appears to increase the
toxicity of the soil/waste system when compared to untreated soils.
  The protocols used in the Western Processing stabilization study
have been adapted  for use as  general treatability protocols for
stabilization  at other Superfund sites.

INTRODUCTION
  Following  publication  of the Western  Processing Feasibility
Study (1) in  1985, soil  stabilization was identified as a potential
remedial measure. Inadequate data were available to evaluate this
alternative. A large-scale bench study was designed as a supple-
ment to the feasibility study to estimate the leaching characteristics
of stabilized  Western Processing soils.
  Seven tasks comprised the treatability study. These include:

  Develop a treatability protocol for soil stabilization
  Identify heavily contaminated surface soil
  Collect representative soil samples
  Characterize representative soil
  Stabilize/vitrify soil
  Characterize stabilized/vitrified soil
  Evaluate environmental impact

EXPERIMENTAL APPROACH

Treatability Protocol for Soil  Stabilization
  A treatability protocol was  needed to answer two  questions:
(1)  will soil  stabilization enable  site  remediation goals  to be
achieved, and (2) how much will stabilization improve the condi-
tion of site soils compared to untreated soils? The U.S. EPA's soil
                                                         stabilization/fixation handbook provided guidance on how these
                                                         questions might be answered G>.
                                                           Stabilized soils can be evaluated for several physical and chemical
                                                         characteristics. They also can be evaluated for biotoxicity. The
                                                         former is the subject of the stabilization/fixation handbook; the
                                                         latter is the subject  of contemporary, promising research.
                                                           The potential methods to evaluate whether stabilization could
                                                         meet site remediation goals include demonstrating that full-strength
                                                         leachates meet all health and environmental goals, risk assessment
                                                         procedures based on simplified methods of pas, sing leachates to
                                                         receptors or formal transport modeling. Since a calibrated three-
                                                         dimensional groundwater model was available for the site, modeling
                                                         was chosen as the method to demonstrate probable conformance
                                                         with site remediation goals.
                                                           The method used  to compare untreated to stabilized soils was
                                                         to subject each to the identical set of test procedures. Because soil
                                                         systems are expected to exhibit great variability, multiple replication
                                                         of all tests so that reasonable statistics can be generated is desirable.
                                                           Combining multiple  replications of all tests  for each of the
                                                         physical, chemical and biological candidate tests  suggests two
                                                         requirements: (1) a considerable volume of soil  is required, and
                                                         (2) simplification as to  procedures or chemical analytes, if pos-
                                                         sible, is necessary. That is, if a system-controlling chemical can
                                                         be identified, then most analyses should  be  limited to it.
                                                           Chemical tests deemed important include full characterization
                                                         of untreated and stabilized soils, and leaching by a variety of proce-
                                                         dures. The  EP  toxicity  test and the Toxicity and  Toxicity
                                                         Characteristic Leaching  Procedure (TCLP) were used to determine
                                                         regulatory status. The Solid Waste Leaching Procedure (SWLP)
                                                         was used to provide engineering estimates of leachates which might
                                                         be expected  from this site.
                                                           Physical tests included particle size/clay fraction, organic con-
                                                         tent, permeability, unconfined compressive strength and durability
                                                         as evaluated by wet/dry stressing (this site is subject to wetting and
                                                         drying but not to freezing and thawing).
                                                           Biotoxicity tests which are easily adapted for use in soil systems
                                                         include a series of aquatic and terrestial bioassessment procedures
                                                         developed by the U.S. EPA's Environmental Research Laboratory
                                                         (3. 4)

                                                           The quantities of  material needed for each test were calculated,
                                                         multiplied by  (nominally) three replicates per test and added to
                                                         an arbitrary reserve  to yield a material requirement of 1400 Ib. of
                                                         contaminated  soil.
                                                           The final issue resolved in the protocol  was the  method of
                                                         stabilizing and vitrifying the soil. The alternatives included the use
198
TREATMENT

-------
of vendors to demonstrate the capabilities of their products or the
use of generic techniques. The former captures the expertise of
industry, while the latter produces data which might be more easily
incorporated in a public sector construction bid specification. An
additional consideration with the vendor approach is the level to
which  the details of the vendor's process or  product  must be
understood.
  Seeking to achieve the best possible result, this protocol used
the vendor approach. To keep the protocol as simple as possible,
each vendor/product was treated as a "black box," (i.e., vendors
were provided with soil and were only required to deliver final
product to various laboratories for testing).

Identify Heavily Contaminated Surface  Soil
  Fourteen hundred pounds of contaminated soil were required
to complete this treatability study. A single sample representing
the most heavily contaminated surface soil was chosen to repre-
sent the most difficult conditions likely to be encountered in a full
scale project. This choice excluded other possible samples such as
the "most representative" surface soil or the "most difficult" soil
matrix.
  The most heavily contaminated surface soils were identified by
reviewing all available soil data(5) and by obtaining a series of sur-
face soil samples from a number  of locations on the  site. These
samples were analyzed using screening techniques for polynuclear
aromatic  materials, thought  to be  the most prevalent organics
remaining at the surface (6).
  Fifteen soil sampling stations were located at Western Processing
by  imposing a randomly oriented equilateral triangle net on the
site. Each triangle defined  an area of 0.3 acre. Eleven of the 15
stations were located on soil,  the remaining four were located on
asphalt or concrete. Samples from the eleven  soil stations were
analyzed for polynuclear aromatic  hydrocarbons (PAH) using field
screening methods. Table 1 summarizes all screening data. Based
on these data, contamination maps from the remedial investigation
and secondary factors such as access, workability, presence of non-
native materials and health and safety, station WPIS-9 was selected
as  the location for the stabilization evaluation.
                           Table 1
                     Field Screening Data
            (sum of priority PAH's concentration pg/1
   Station

   WPIS -  04
   WPIS -  05
   WPIS -  06
   WPIS -  07
   WPIS -  09
   WPIS -  10
   WPIS -  11
   WPIS -  12
   WPIS -  13
   WPIS -  14
   WPIS -  15
Total  PAH

     8,000
     9,000
    12,000
    44,000
    91,000
    71,000
    32,000
    34,000
    15,000
   740,000
   100,000
 Collect Representative Soil Samples
   A sampling plan to develop a 1400 Ib. homogeneous sample was
 prepared.  It established a 4-ft. by 6-ft. rectangle centered over
 station WPIS-9. The soil crust, approximately 2 to 3 in. deep was
 skimmed,  photographed and discarded. The next 12 to 15 in. of
 soil were hand excavated in stages into a specially prepared sample
 container. The container was a sheet of standard plywood with
 12 in  batter boards. Soil was transferred to the sample container
 in several  small lots. As each lot was completed, the sample box
 was divided into  four  cells and all soil was quartered and re-
 quartered. All foreign materials greater than 1.5 in. in diameter
 were removed and photographed. When the sample box was filled,
it was quartered multiple times. (Based on available soil data it
was suspected that volatile materials were not present at WPIS-9;
therefore, no effort was made to retain volatile materials).
  The sample container was divided in 8 equal 4 ft2 cells which
were consecutively numbered. The soil was transferred approx-
imately 4 Ib. at a time into 69 5-gal containers, keeping the final
weight of container and soil as close to but less  than 25 Ib. The
bucket into which a portion of soil was placed was determined by
a random number table; each bucket required five to six portions
to fill.
  A second random number table was used to divide the 69 buckets
between the stabilization and vitrification vendors, analytical
laboratories and a reserve.  Each vendor required from 2 to 7
buckets of soil.
  The success of the treatability study depended  on each partici-
pant receiving an essentially equivalent portion of soil. Considerable
planning and effort ensured that this was the case. The following
elements of the stabilization protocol are judged to be of critical
importance. The vendors were advised to:
  Mix soil thoroughly according to a formal plan
  Apply the mixing technique for longer than deemed necessary
  (i.e., determine what is necessary and double  it)
  Place the treatability sample into multiple identical
  containers using proper random fill techniques
  Allocate the containers to various study purposes using proper
  random techniques

  A considerable investment is required  to develop treatment
samples in this manner. They are properly viewed as Site Reference
Samples (SRSs). Being a reference sample, a sufficient  reserve
should be developed for other site purposes.

Characterize Representative  Soil
  Untreated soils were characterized for physical properties by the
U.S. Army Corps of Engineers North Pacific Division (NPD)
Materials Laboratory, Troutdale, Oregon, for biotoxicity by the
U.S. EPA Environmental Research Laboratory, Hazardous Waste
and  Water Branch, Corvallis, Oregon,  and for chemical and
leaching characteristics by the U.S. EPA Region 10 Environmental
Laboratory.
  Organic content of the soil was less than 5%. Less than 2% by
weight of soil was gravel (grain size > 5 mm). The silt/clay fraction
(grain size <0.075 mm) was 47%. Maximum dry density was
102.9 lb/ft2 at  an optimum water content of 17.9%.
  Permeabilities  were  determined  on  samples  remolded  to
maximum dry density at optimum water content, 85% and 95%
of maximum density on the wet side of optimum, and 85% and
95% on the dry side of optimum. Tests were made under pressures
     100 -

     90 -

     to -

     70 -

     60 -

     50 -

     40 -

     30 -

     20 -

     10 -

      0 -
                           0      *


                        D  p — 0 TSF
                           Son Uol.tur. In X
                           f - 0.5 TSF
                                                                     p - 1 TSF
                                               Figure 1
                                       Untreated Soil Permeability
                                                                                                          TREATMENT    199

-------
corresponding to 0, 12 and 20 feet overburden (i.e., 0, 0.5 and
1 ton/ft2). Data are summarized as Figure 1. Noteworthy is the
10-6  cm/sec permeability  achievable by bringing  the  soil to
maximum density at optimum water content.
  Biological characterization of the soil included both aquatic and
terrestial bioassays.  Three aquatic  assays, Selenastrum capri-
cornutum (an algal procedure), Daphnia magna and Microtox (a
commercial  bacterial procedure) were selected.  Three terrestial
procedures were also selected including root elongation, Neubauer
(seed germination), and Eisenia andrei (an earthworm procedure).
Results are expressed as EC50 or LC50 values, the percentage of
sample which elicits a 50% response (such as mortality) in the test
organisms; the lower the EC50 or LC50, the more toxic the sample.
The following scale was used  to interpret bioassessment data:
                           Table 2
             Characterization of Bioassessment Data
   EC50  ฃ LC50
   >10t,<20t
   >20*,<75%
   >75%  or inhibition
   no effect (NE)
                       Toxicity Label

                       extremely high
                       high
                       moderate
                       low
                       none
   As evaluated by D. magna, the toxicity is moderate. As evaluated
by S. capricomutum, the eluate soil toxicity is high to extremely
high.
   Chemical characterization of test soil included complete priority
pollutant analyses and leachate testing by EP Toxicity, TCLP and
SWLP. In contrast to the field screening data and previous soil
samples in the vicinity of WPIS - 9, no priority pollutant organics
were present  in a  leachable form. Most leachate testing was
therefore  limited to  seven  key metals:  zinc, lead, cadmium,
chromium, copper, nickel and barium. Table 3 presents leaching
results for the test soil:
                           Table 3
                Leaching Procedure Comparison
                   (all data in ug/l, or ug/kg)
  Contaminant

       Zn
       Cd
       Pb
       Cr
       Cu
       Ni
       Ba
    Total

5,000,000
    27,000
1,200,000
   170,000
   110,000
    30,000
    70,000
EP  Toxicity

  68,000
      360
      530

        47
        67
      450
  TCLP

68,000
     20
 5,100
     45
    170
     73
    980
   SWLP leachates differ from EP Toxicity and TCLP leachate
 in that background site water or distilled/deionized water is used
 as the eluent rather than an acid solution. Each SWLP test results
 in  sequential leachates being generated. SWLP results  for
 chromium,  lead, zinc and cadmium are presented as Figures 2
 through 5. Each leachate within an SWLP sequence is interpreted
 to represent leaching behavior at a later time than the cycle before.
 For diffusion-limited leaching, the expected  trend is for  lesser
 amounts of  contaminant to be released with increasing time. These
 data show increasing rates of release with increasing time, a finding
 which is unexplained but which may be related to the speciation
 of the contaminants, the loss of modest soil buffering capacity or
 leaching controlled by dissolution.
   Ambient  water quality criteria are limiting for metals at  the
 Western Processing site. These criteria are plotted on Figures 2
 through 5 against SWLP leachates to provide  some indication of
 the  significance  of leachates to site remediation needs.
                                                     26

                                                     14 -

                                                     a -

                                                     20 -

                                                     II -

                                                     It -

                                                     14 -

                                                     U -

                                                     10 -
                                                      4 -

                                                      2 -

                                                      0
                                                   mojdmum aquattc criteria
                                                                                                      21 ti.ir •quctk if* critarta
                                                            Figure 2
                                                  Solid Watte Leaching Procedure
                                                                            Figure 3
                                                                  Solid Waste Leaching Procedure
900 -


too -


700 -


•00 -


soo -


400 -


300 -


200 -


100 -


  0 -
                                                                  2< hซur aqtMh Oh critwta
                                                                        Mcond              th

                                                                        IP Crtractlo* s^inno tar 2>c
                                                                             Figure 4
                                                                   Solid Waste Leaching Procedure
200    TREATMENT

-------
                                    maximum aquatic criteria
                  24—hour aquatic life criteria
                        ••cond              third

                         SWLP •xtrocU of cadmium
                           Figure 5
                Solid Waste Leaching Procedure
                                                                   density of their product. In developing a construction specifica-
                                                                   tion for stabilized soils where permeability is a concern, a density
                                                                   target may be needed.
                                                                     The bioassessment  data suggest  a  degradation  in soil
                                                                   characteristics compared to the untreated soil. Using the ter-
                                                                   minology presented in Table 2, Tables 4 and 5 present the results
                                                                   of the bioassessment  testing.
                                                                                              Table 4
                                                                                     Aquatic Bioassessment Tests
                                                                                          Of Treated Soils
                                                                                         (degree of toxicity)
 Vendor

 raw coil

 Chlyoda
 Calweld
 Soiltech
   (Firestone)

 Battelle
Selenastrua

extremely high

high
high

extremely high

none
 Daphnia

 moderate

 none
 low

 high
                                                                                                                      Microtox

                                                                                                                      nona
Moderate
moderate
moderate

none
Stabilize/Vitrify Soil
  Three commercial stabilization vendors volunteered to participate
in this treatability study. Each was issued by random lot seven con-
tainers of contaminated soil. These were recombined, stabilized
and cast into 14 standard 6- by 12-in concrete test cylinders. (Large
cylinders were chosen to minimize the effect of large particles on
the cylinder characteristics and to encourage the use of mixing
techniques more similar to those  which might be encountered in
the field).  The actual stabilization process was not supervised.
Chemical compositional analyses  were used to ensure that the soil
samples were not altered prior to  stabilization. All cylinders were
shipped to the NPD Materials Laboratory for curing and sample
preparation.
  Cooperating vendors included the Chiyoda Company, Soiltech
(now Firestone Resources, Inc.)  and Calweld/ATW.
  Two containers of soil were provided to the vitrification vendor,
Battelle Pacific Northwest Laboratories. The soil was vitrified with
clean matrix soil into a monolith  weighing approximately 600 Ib.
This, too,  was forwarded to the  NPD Materials Laboratory for
processing.

Characterize Stabilized/Vitrified Soil
   Stabilized and vitrified soils were subjected with few exceptions
to the same tests as the untreated soil. Differences included:
• Permeabilities  were not developed for the vitrified soil
   (an appropriate procedure could not be identified)
• EP Toxicity was not used with any treated materials
• The chemically stabilized soils were subjected to stressing by the
   wet  cycle/dry cycle test

   All chemical stabilization vendors achieved unconfined com-
pressive strengths  between  110  and  170 lb/in2.  The  Chiyoda
Company product developed the greatest strength. The unconfined
compressive strength of the vitrified soil exceed 30,000 lb/in2.
   Two chemical  stabilization vendors completed. 12 full cycles of
wet/dry testing with weight losses of less than 5% and volume
change of less than  1%. The third vendor,  Soiltech (Firestone
Resources), failed at 10 1/2 to 11  1/2 cycles, losing all weight. (The
standard wet/dry test was modified to eliminate the wire brushing
step on generally weak materials).
   Permeabilities of the chemically stabilized soils varied less with
overburden pressure than did  the untreated  soils. Excluding
outliers, the stabilized permeabilities ranged from 0.3 to  30 x 10~6
cm/sec.' These values are roughly equivalent to the permeability
exhibited  by the untreated soil when at optimum  water con-
tent/maximum density. All stabilized soils were poured and rodded
into their cylinders; none of the vendors attempted to improve the
                          Table 5
          Terrestial Bioassessment Tests of Treated Soils
                      (degree of toxicity)
                  Root            Seed
Vendor

raw soil

Chiyoda
Calweld
Soiltech
 (Firestone)


Battelle
   Elongation

   none

   low
   low

   low


   low
Germination

none

moderate
moderate

extremely
 high
 Earthworm

 none

 moderate
 high

 extremely
  high
  The vitrified product clearly improved the condition of the soil
as measured by the aquatic bioassays. Two of the three chemical
stabilization vendors slightly improved to slightly degraded the con-
dition of the soil. One vendor, Soiltech (Firestone), failed to im-
prove the soil as indicated by each of the three aquatic bioassays.
  The vitrified soil maintains the soil in a no toxicity condition.
This is in sharp contrast to all chemically stabilized products which
increased slightly to greatly toxicity as measured by the terrestial
bioassays. Soiltech (Firestone), for example, increased observed
toxicity for the Neubauer (seed germination) and E. andrei (earth-
worm) assays from no observed effect with untreated soil to EC50's
of 4ฐ7o and 9% respectively. A mechanism to explain this increased
toxicity was not discovered. However, the moderately elevated pH
of the stabilized material was demonstrated in subsequent tests not
to be a factor.
  Chemical characterization of the stabilized and vitrified products
consisted largely of TCLP and  SWLP leaching tests. Figures 6
through 9 summarize the leaching results achieved by the chemical
stabilization vendors. These Figures compare directly to Figures
2 through 5 for untreated soils.
  Several items in these data are of interest. Leaching rates have
been decreased significantly. As measured by the TCLP procedure
for zinc (the dominant contaminant on the site), the most effective
vendor, Calweld/ATW,  reduced  the leachate  by greater  than
99.99 + % to a value of 7  ug/1. The SWLP data tend to decrease
with each cycle, compared  to untreated soil where it increased.
   SWLP testing included monolithic samples and samples pul-
verized to the  TCLP standard of  <9.5 mm.  The  tumbling
apparatus used in both tests was the 30 rpm end-for-end tumbler
specified in the TCLP procedure. This effectively reduced the size
of all samples to a fine  powder within hours. The monolithic
                                                                                                           TREATMENT    201

-------
leaching performance as determined in these tests, therefore, did
not differ significantly from the pulverized sample leaching data.
  The vitrified sample included nearly 50 Ib. of contaminated soil
melted into nearly 600 Ib., or a ratio of approximately 1:10, con-
taminated soil to inert soil. Table 6 compares vitrification leaching
data to leaching data from untreated soils and the most effective
chemical  stabilizer.

                           Table 6
                 Leaching Behavior of Stabilized
                      And Vitrified  Soils
                         (zinc In ug/l)
   Leach Test

   TCLP

   SWLP, extract 1
   SWLP, extract 2
   SWLP, extract 3
   SWLP, extract 4
Untreated

 60,000

    190
    450
    530
    • 20
Vitrified

  230

   66
  200
  220
  310
Stabilited

    7

   29
   15
   1ซ
   14
     100
      90 -


      80 -


      70 -


      60 -


      SO -


      40 -


      30 -


      20 -


      10 -

                     EXTRACT ป1
                                EXTRACT |2
                                           EXTRACT 13
                       TCLP and SWLP Uachlna fncuurt__
                          l\\l Chtyoda        E2Z)
                            Figure 6
                  Stabilized Soils for Chromium
                                                                                                Figure 8
                                                                                         Stabilized Soils for Zinc
                                            TCLP


                                         Cr7\  CoMd
                                                                   EXTRACT t

                                                                      TCU" cป
                                                                                                     CXTRACT f 2
                                                                                                                OCTRACT H
                                                                                                                            EXTRACTS

90 -
80 -
!
= 70 -
i
i
! tO -
' SO -
!
' 30-
, X H
20 -
10 -
0 —








I
//\ wfa //

R^
I

SWLP
\^ SWLP SWLP
p-7-VN^ f-TT-fvsfe^l KN^
TCLP EXTRACT ft EXTRACT fl EXTRACT |3 EXTRACT f<
TCLP oniSWLP Uochlng Pme.aur, 	
I/ /I Oalw.W IVXl CWyodo K7^1 lolll.eh
                            Figure 7
                     Stabilized Soils for Lead
                                                                                                 Figure 9
                                                                                        Stabilized Soils for Cadmium
                                                 The vitrified product as evaluated with these standard leaching
                                               tests did not perform well. The reason for this is not known. It
                                               may be that  the  nature  of the tests may be inappropriate for
                                               monolithic,  vitrified  masses,  or  vitrification might not be as
                                               effective as chemical  stabilization for simple metal systems.

                                               Evaluate Environmental  Impact
                                                 The  performance  tests provided site-specific  geochemical,
                                               sorption, hydraulic conductivity and leachate data that were used
                                               to evaluate a soil stabilization remedial action for the Western Pro-
                                               cessing site. Other remedial actions considered for the site included
                                               no action, source removal, pump and treat, and capping.  The
                                               relative performance of these remedial actions was evaluated using
                                               the three-dimensional Coupled Fluid Energy and Solute Transport
                                               (CFEST) code (6). The flow simulation model was developed and
                                               calibrated to measured  potentials and flow  rates on site.  The
                                               transport  simulation model was used to predict the movement of
                                               three constituents: trichloroethylene (TCE), zinc and cadmium. The
                                               transport  model was calibrated to estimate total constituent mass
                                               in the system, measured  ground.water concentrations of the site
202     TREATMENT

-------
and estimated mass flux to the nearby creek.
  The site-specific performance test data were used to establish
the following model input parameters for  the soil stabilization
remedial action simulations: (1) horizontal  and vertical conduc-
tivity, (2) recharge, (3) leachate concentration (short-term strength)
and (4) saturated zone concentrations. It was assumed that the top
20 ft. of soil was stabilized over the entire 13, acre site.
  Performance testing indicated that the hydraulic conductivity
is reduced approximately one order of magnitude as a result of
stabilization. Conductivities in the model were, therefore, reduc-
ed one order of magnitude throughout the stabilized zone. The
surface recharge due to precipitation was similarly reduced by a
factor of 10 in the model.
  Leach tests of the stabilized soils showed a one to two order of
magnitude reduction in the leachate concentration for the metals.
Thus, the leachate concentration in the unsaturated zone (source
term of the model) and the saturated zone concentrations in the
stabilized area were  reduced by a factor  of 10 for zinc and
cadmium. The soil sample obtained from the site did not contain
TCE, It was therefore not possible to measure the effect of stabiliza-
tion on TCE concentrations. The unsaturated and saturated zone
TCE concentrations were not changed in the model based on these
findings.
  When simulating contaminant transport with the CFEST code,
it is necessary to estimate a distribution coefficient (Kd) for each
constituent being modeled and the duration  of time until the con-
stituent completely leaches from the soil source. Performance
testing provided site-specific Kd's for zinc and cadmium. Based
on the measured Kd's, the initial inventory and the recharge rate,
the duration  of time until leaching is completed (release time of
source term) was estimated.
  The stabilization action very effectively controlled migration of
the metals. In the no-action simulations, both zinc and cadmium
migrated to Mill Creek at unacceptable levels. After stabilization,
these metals became bound upon the stabilized soil and the resulting
creek  concentrations were  well  below acceptable levels. The
stabilization  action was not as effective for TCE. TCE is more
mobile than the metals, and the majority of the  TCE plume had
already migrated out of the stabilized area before the action was
implemented in  the model. If the entire soil mass contaminated
with TCE was stabilized, the action would be much more effective.
  Based on the comparison to other remedial actions, the stabiliza-
tion action appears to most effectively control the highly retarded
contaminants (metals), while the pump-and-treat  plus source-
removal actions appear to most effectively control the more mobile
contaminants such as TCE.

CONCLUSION
  Chemical stabilization and vitrification are effective remedial
measures  for  soil  systems  contaminated with heavy metals.
Reasonable  treatability studies  can demonstrate their efficacy.
Treatability studies should consider the physical, chemical and
biological properties of the soil systems in question.

ACKNOWLEDGEMENTS
  The support of the  U.S. EPA Region  10 Environmental
Laboratory  in characterizing materials and in developing the
capability to perform numerous leaching tests at a high throughput
is  acknowledged.  The U.S.  Army Corps of Engineers  NPD
Materials Laboratory is acknowledged for materials handling and
physical testing. The U.S. EPA Hazardous Waste Engineering
Research Laboratory provided the technical approach. Finally, the
cooperation of the four commercial vendors (Chiyoda, Soiltech,
Calweld/ATW, and Battelle Northwest) is acknowledged for the
professional manner in which they provided product for testing.

REFERENCES

1.  CH2M-HH1, "Feasibility Study for Subsurface, Western Processing,
   Kent, Washington," (2 volumes), March 6, 1985.
2.  U.S. EPA, "Handbook for Stabilization/Solidification of Hazardous
   Wastes," EPA/540/2-86/001, June 1986.
3.  Porcella, D.B., "Protocol for Bioassessment of Hazardous Waste Sites,"
   EPA/600/2-83/054, U.S. EPA,  Corvallis, OR, 1983.
4.  Miller, W.E., Greene, J.C., Peterson, S.A., "Revised Protocol for
   Bioassessment of Hazardous Waste Sites," (in preparation), U.S. EPA,
   Hazardous Waste and Water Branch, Corvallis, OR, 1987.
5.  CH2M-Hill, unpublished data, (Available in DBASE III).
6.  CH2M-HU1,  "Final Remedial Investigation Data Report, Western
   Processing, Kent, Washington," December 17, 1984.
7.  Cupta,  S.K.,  Meyer, P.R., Newbill, C.A., and Cole, C.R., "A
   Multidimensional Finite Element Code for the Analysis of Coupled
   Fluid, Energy, and Solute Transport (CFEST)," PNL 4260, Battelle
   Pacific Northwest Laboratory, 1982.
                                                                                                         TREATMENT     203

-------
                                 Groundwater  Restorations At
                                              McClellan AFB

                                                    Mario  lerardi
                                                 Chief,  Installation
                                               Restoration Program
                                                  McClellan  AFB
                                              Sacramento, California
                                                    Paul Brunner
                                                  Deputy Director
                                           Environmental Management
                                                  McClellan  AFB
                                              Sacramento, California
                                            Edward J, Cichon, Ph.D.
                                                 General Manager
                                               Metcalf  & Eddy, Inc.
                                               Palo Alto,  California
ABSTRACT
  An advanced groundwater  treatment  system  is  operating
successfully as part of the hazardous waste cleanup program at
McClellan Air Force Base. Representing a concerted effort by the
Environmental Management Group at McClellan AFB over the
last 7 years, the treatment system is one of the first of its size and
type for groundwater cleanup in the United States.  Groundwater
contaminated with organic chemicals and  metals can be treated
at a rate  of 1000 gal/min  to produce an  effluent which meets
drinking water standards. The treatment system has passed a com-
prehensive 30-day performance test and is now under continuous
operation by Metcalf & Eddy's hazardous waste site cleanup
division.
  The groundwater remediation system consists of: (1) a series of
complex clay and membrane layers which  form a protective cap
over the contaminated area, (2) a network  of six extraction wells
and a 10,000-ft pipeline which deliver groundwater and (3) a
groundwater treatment facility (GWTF). The processes used to
restore the groundwater to drinking water quality include high tem-
perature air stripping, incineration,  biological degradation and
granular activated carbon adsorption.
  Ensuring  that contaminants would be destroyed  in the GWTF
and not simply transferred to another environmental medium was
a driving force in the selection of treatment technologies. Data will
be presented which reflect successful performance of the GWTF
and obstacles overcome during design and construction will be
discussed.

INTRODUCTION
  Prior to the mid-1970s, improper management and disposal of
hazardous wastes were common among industry and government.
Although unintentional, significant  contamination of soil and
groundwater frequently occurred  as a result of these  unsound
practices.  Organic solvents placed in landfills and impoundments
have migrated to groundwater used for drinking water supplies and
for agriculture. Soils contaminated with metals, petroleum residues,
explosives and other wastes pose threats to public health and the
environment.
  The Air Force, as a large industrial activity and landowner is
responsible for some of the environmental contamination problems.
As a result, the Installation Restoration Program (IRP) was imple-
mented by DOD in 1980 to cleanup health threatening hazardous
waste sites at military installations.  McClellan AFB in Sacramento
has been widely recognized within the Air Force and DOD as a
leader in its management and implementation of the IRP. One of
the most significant achievements under the IRP at McClellan is
the advanced groundwater treatment system which is part of the
remedial action program for Area D.
  Representing a concerted effort by the Directorate of Environ-
mental Management at McClellan AFB over the last 7 years, the
treatment system for Area D is one of the first of its size and type
for groundwater cleanup in the United States. Groundwater con-
taminated with organic chemicals and metals can be treated at a
flow  rate of  1000 gal/min  to produce an effluent which meets
drinking water standards. In January 1987, the treatment system
passed a comprehensive 30-day performance test; and since then
it has been in continuous operation by Metcalf & Eddy's hazardous
waste site cleanup division.

Background
  As mentioned above the Area D groundwater treatment facility
(GWTF) represents pan of the IRP Phase IV effort at McClellan
AFB. The objectives driving the selection of a remedial system for
groundwater decontamination are shown in Table 1. In addition,
                         Tibk 1
         McClellan AFB Groundwater Treatment Facility
                    Treatment Objectives
    Contaminant




Volatile Orgซnics


Non-Volatile Organic*


Ketones

Me to la



VOC Emission*


(D
                                        Effluent Criteria
Belov Analytical
Detection Llnlta
(D
Below Analytical
Detection Llalts  (D

Below 1 ppn

Below Primary
Drinking Water Standards

Zero (2)
(2)
    Typically 0.5 ppb
                                                                  Later defined as 99.99Z DRE
204    CONTAMINATED AQUIFER CONTROL

-------
the system was to be transportable thereby allowing for potential
use at other sites, and was to have a 25-year design life in order
to accommodate an extended remedial period at Area D.
  Groundwater quality data obtained from monitoring  wells
installed in and around Area D as part of the remedial investigation
were used to screen and select the groundwater treatment process.
These data, summarized in Table 2 for the major contaminants,
were used to prepare the entire design, construction and contract
operations bid.
  The GWTF designed and constructed at McClellan AFB had to
satisfy two other key criteria in addition to those delineated in Ta-
ble 1. First, the GWTF had to accommodate varying flowrates from
the Area D extraction wells, ranging from 250 to 1,000 gal/min.
Second, the treatment system had to be flexible enough to handle
contaminants  from other areas on the base.
                            Table 2
          Groundwater Contaminant Concentrations Area D
                                                               Table 3
                                                    Potential Treatment Technologies
         CONTAMINANT
  1,1-Dichloroethylene
  1,1,1-Tr ichloroethane
  Trichloroethylene
  Methylene Chloride
  1,1-Dichloroethane
  Trans 1,2-dichloroethylene
  Benzene
  Vinyl Chloride
  Toluene
  Tetrachloroethylene
  Dichlorobenzenes
  Phenols
  Acetone
  MEK
  MIBK

  Chromium
  Nickel
  Zinc
  Lead
  Selenium
  Cadmium
   MAXIMUM
CONCENTRATION
     PPb
    63,000
    12,000
    11,000
     5,000
      250
      200
      680
     2,500
       80
       70
      170
      500
    35,000
    24,900
     3,700

      120
      100
       73
       93
       49
       12
   AVERAGE
CONCENTRATION
     PPb
    11,700
     3,600
     3,000
     3,200
      140
      120
      350

       47
       30
       83
      460
    22,400
     8,500
     2,500

       93
       68
       44
       51
       49
       11
   The general approach taken by Metcalf & Eddy to implement
 the remedial actions consisted of the following steps. A 4-week
 feasibility study of the various remedial alternatives was conducted
 resulting in the identification of the most cost-effective ground-
 water treatment alternative. This feasibility study was followed by
 a detailed design phase, construction of the treatment system,
 preparation of an operations and maintenance manual, start-up
 of the GWTF and long-term contract operations.
   A theme prevalent in the various phases of this project was to
 avoid producing residuals at the GWTF which would require treat-
 ment and/or disposal off of the base. The groundwater treatment
 system described in the following sections demonstrates how this
 goal of ultimate on-site treatment was accomplished.

 PROCESS SELECTION
   The GWTF feasibility study and detailed design were accom-
 plished without having to go through pilot scale testing. Process
 performance data from previous full-scale groundwater remedia-
 tion systems designed, constructed and/or operated by Metcalf &
 Eddy were utilized to select  the system for  McClellan.
   Contaminants  from Area D were divided into  four general
 categories based on their physical and chemical characteristics; vola-
 tile organics, non-volatile organics, ketones and metals. Applica-
 ble  treatment technologies  were identified  for each type of
 contaminant (Table 3) followed by integration of separate tech-
.nologies into process alternatives.
                        CONTAMINANTS



                  1.  Volatile Organics


                  2.  Non-Volatile  Organics


                  3.  Ketones


                  4.  Metals


                  5.  VOC Emissions
                                                                                       TECHNOLOGY
                                                                                     Air-Stripping
                                                                                     GAC

                                                                                     GAC
                                                                                     Biological

                                                                                     Biological
                                                                                     GAC

                                                                                     Ion-Exchange
                                                                                     GAC

                                                                                     GAC
                                                                                     Thermal
                                                                                     Catalytic
  To remove  volatile organics,  Metcalf & Eddy's proprietary
VOLSTRIP program was utilized to  simulate the air stripping
process under various organic loadings, flow rates, air-water ratios
and  temperatures. The initial results indicated that capital and
operating costs could be minimized by employing a high tempera-
ture  air-stripping process to remove volatile organics to concen-
trations less than analytical detection limits. In order to meet the
VOC emissions criteria for the GWTF, several technologies were
evaluated for treatment of the air-stripper off-gases. These tech-
nologies  included vapor phase GAC,  incineration and catalytic
destruction. The catalytic destruction option was eliminated from
further  consideration due to the high probability for catalyst
poisoning. At  the beginning of the evaluation of off-gas control
technologies, existing data indicated vinyl chloride concentrations
in the groundwater to be in the range of 30-90 jtg/1. These levels
could be accommodated by vapor phase GAC. However, additional
groundwater quality data generated during the project showed vinyl
chloride concentrations as high as 6,500 /tg/1, thereby precluding
vapor-phase GAC from further consideration as a VOC emission
control technology.
  The incineration system was chosen as  the most cost-effective
means of treating air stripper off-gases prior to discharge to the
atmosphere. Advantage was taken of waste heat generated by the
incinerator for use in heating influent groundwater for  the high
temperature air-stripping process. In keeping with the theme to
minimize residuals generated by the GWTF, a caustic scrubber was
selected to neutralize acid gases in the incinerator exhaust. The high
temperature air-stripping option with off-gas incineration for treat-
ment of volatile organics was economically more favorable than
a liquid phase GAC  system.
  Groundwater leaving the air-stripper would still contain ketones,
non-volatile organics and metals. Biological treatment and GAC
were identified as the most viable alternatives for treatment of
ketones. Biological treatment was considered due to the biode-
gradable characteristics of acetone, MEK and MIBK; GAC was
included for evaluation due to its anticipated use in removing the
non-volatile organics.
  From previous experience, it was recognized that some initial
adsorption of ketones could occur on the GAC. Given the expected
concentrations in the groundwater from Area D, it was predicted
that  breakthrough in a 60,000 Ib. GAC system would occur wi-
thin  2 months. This breakthrough period was later confirmed dur-
ing full-scale tests at the GWTF.
  Several biological treatment processes were considered for ketone
removal including activated sludge and various fixed-film reactors.
For the Area D GWTF, a fixed film aerobic biological treatment
process was chosen over activated sludge  due to a more compact
reactor design, lower electrical power requirements, relatively low
labor requirements and simplified liquid/solids separation. The
advantages of selecting the fixed-film biosystem over GAC for

            CONTAMINATED AQUIFER CONTROL    205

-------
ketone removal were (1) ketones are destroyed and not merely
separated from the groundwater, (2) phenols and phthalates are
removed to some extent, reducing the downstream carbon usage
rate for removal of non-volatile, non-biodegradable organics and
(3) very low operating costs compared to GAC.
  Effluent  from the biosystem would still  contain recalcitrant
organics plus trace levels of metals. GAC was identified as the most
viable process to remove the remaining organics. The GAC system
was also designed with backwashing capabilities in order to handle
any solids carryover from the biosystem. The low concentrations
of metals in the groundwater eliminated conventional precipita-
tion processes from further consideration as treatment alternatives
for achieving  the metal  discharge  limitations. A  comparison
between GAC, which has a relatively small capacity for adsorbing
metal complexes and conventional ion exchange processes showed
the GAC option to be better in terms of residuals  production,
capital costs and operating cost.
  In summary the processes chosen for the detailed design and con-
struction of the Area D GWTF at McClellan AFB are shown in
Figure 1. Groundwater (A) from a network of extraction wells at
Area D is conveyed through a 9,600-ft above-ground pipeline to
the influent storage tank (B). The raw groundwater is heated by
a series of primary  (C)  and secondary  (D) water/water  heat
exchangers and an air/water heat exchanger  (E). Heated ground-
water enters the top of the air-stripping tower (J) where it is met
by a countercurrent  stream of air produced by variable speed
blowers (K). Groundwater minus the volatile organics is pumped
to the fixed-film biosystem (L) where removal of ketones occurs.
Finally, effluent from the biosystem is pumped though a GAC
system (M) consisting of three parallel trains of two 10,000 Ib vessels
each. Effluent of drinking water quality (N) is discharge to Magpie
Creek.
  Off-gases from the air stripper enter an air/air heat exchanger
(G) for pre-heating prior to combustion of the VOC-laden air-
stream in the incinerator (F). The incinerator is designed to operate
between 1 SOOT and 2300 ฐF with a two-second residence time. Hot
gases from the incinerator transfer heat to the incoming ground-
water through the air/water heat exchanger (E) before passing
through the air/air heat exchanger and into the caustic scrubber
(H).
                       Figure 1

REMEDIAL ACTION
  Construction of the GWTF followed completion of the detailed
remedial design. Several aspects of the GWTF construction are
worth noting. First, the GWTF was located away from Area D

206     CONTAMINATED AQUIFER CONTROL
in a location known as Area C. Thii decision to use this location
was based on (1) the high probability of having to decontaminate
groundwater  from Area C as put of future remedial actions, (2)
the central location of Area C with respect to other areas on base
which may require remedial actions and (3) the proximity of Area
C to the Industrial  Waste Treatment Plant (IWTP) should
temporary diversion of groundwater be required due to an upset
at the GWTF.
   Second, since the GWTF would  handle highly contaminated
water, the plant site was designed and constructed to prevent any
potential leaks or spills from contaminating the surrounding soils
and shallow groundwater. Construction features included exca-
vating the GWTF site at Area C to a depth  of approximately 6
ft, backfilling with a soil/clay/lime mixture, placement of a high
density polyethylene liner system and construction of sumps and
associated laterals.
   In the event that a spill  or leak occurs within the GWTF. the
subsurface liner and collection system will prevent  migration of
contaminants into the surrounding soil routing any spills back to
the headworks of the GWTF.
   Following construction of the GWTF, Metcalf & Eddy initiated
a multi-step startup program for the plant. During the first phase,
the GWTF operated on potable water for hydraulic and mechani-
cal testing and the incinerator was fired up to condition the refac-
tory and confirm control point settings. Subsequent to the potable
water shakedown phase, a 24-hr verification  test  was conducted
with contaminated groundwater from Area D. This  test included
comprehensive sampling and analysis of groundwater at various
locations throughout the GWTF in order to confirm  that removal
efficiencies achieved were in accordance with the design criteria.
   Upon completion of the verification test, the GWTF was shut-
down until analytical test data had been received  and evaluated.
The test results signified successful treatment of the Area D ground-
water and allowed the 30-day performance  test  to begin. This
30-day test included daily monitoring of process performance, but
more importantly was developed to assure that the GWTF could
be operated continuously in accordance with the O&M manual and
discharge criteria. The results of the 30-day performance test are
summarized in Table 4 for major contaminants. Volatile organics,
ketones and metals were removed to below their respective effluent
criteria. For the volatile organics, removal to concentrations less
than 0.5 pg/1 occurred entirely  across the high temperature air
stripping process in accordance with the design criteria. Complete
removal of VOCs by the air-stripper prevented release  of VOCs
in the biosystem and unnecessary loading of VOCs onto the GAC.

                           TซMc4
                 GWTF 30 Day PtrforauuK* Tot
           Dili Summary for Major Contaminants
                                                                                                At!
                                                                                                imam
                                                                    I, l-01cklorathylซM
                                                                    l,l,l-eutUoroซthuป
                                                                    Triehl*roซthylmซ
                                                                    1,1-OicUarMttaM
                                                                    Trtu-1 , 2-OlchloraUyl*
                                                                    Ttnyl ChlorU*
                                                                    ToluซM
                                                                    Ae*tซM

                                                                    HIM

                                                                    llctol
                                1,200
                                  773
                                  •23
                                  3tO
                                2.110
                                  270
                                  JM
                                  4*0
                                21,000
                                S.400
                                2.300

                                0.013
                                0.010
                                                                                                                     40.
40.
40.
•0.
  .13
  .07

  .010
  .01*
  During the 30-day performance test,  the biosystem was not
utilized in order to conduct full-scale tests on the removal of ketones
across the GAC system. It was expected that the high temperature
air-stripping process would reduce influent ketone concentrations

-------
by approximately 30%. Data were collected to ascertain the ability
of GAC to remove the remaining 70% under full-scale operating
conditions.  As shown in  Table  5 and  consistent with process
simulations performed during the feasibility study, acetone break-
through occurred during the third week. These results confirmed
the need to implement the fixed-film bioreactor process for effec-
tive ketone removal.

                          Table 5
             Acetone Concentrations in GAC Effluent
                                        AVERAGE
                                        ACETONE
                                   CONCENTRATION
                                   	PPb
WEEK

   1
   2
   3
   4
                                           0.48
                                           0.71
                                           4.1
                                           4.1
CONCLUSION
  Since completion of the 30-day performance test in January 1987,
the GWTF has been operated by Metcalf & Eddy and successfully
treating contaminated groundwater from Area D. The operating
data from the GWTF have demonstrated that:

• Proper integration of proven technologies can effectively treat
  a complex mixture of organic chemicals  in groundwater to
  below analytical detection limits
• Successful groundwater remediation systems  can be designed,
  constructed and operated without having to resort to costly and
  time-consuming  pilot studies
• Cost-effective remedial actions can be implemented without
  generating residuals requiring off-site treatment and/or disposal

ACKNOWLEDGEMENTS
  The authors would like to thank Col. Thomas J. Lawell, Direc-
tor of Environmental  Management at McClellan AFB for his
guidance  and support throughout this project. In addition, the
authors are grateful to Mr. Fred Edgecomb, Plant Superintendent
for Metcalf & Eddy, and his staff for the valuable data obtained
from the  GWTF operations.
                                                                           CONTAMINATED AQUIFER CONTROL    207

-------
       'Decay  Theory" Biological Treatment  for  Low-Level  Organic
                Contaminated Groundwater and  Industrial  Waste

                                                  Kevin M.  Sullivan
                                                    DETOX, Inc.
                                                  Ithaca,  New York
                                                 George J. Skladany
                                                    DETOX, Inc.
                                            Newport Beach, California
ABSTRACT
  A "decay mode" biological submerged fixed-film reactor has
been designed to treat groundwater and industrial waste waters con-
taining less than 50 mg/1 total influent organics. The ability of this
reactor to successfully treat organic concentrations below those
generally thought to be amenable to biological treatment is based
upon the application of microbial "decay" rather than "growth,"
processes. A healthy biofilm initially grown at high organic con-
centrations  within the reactor  is able to continue  scavenging
organics from water after it has been switched to a feed consisting
of low (less than 50 mg/1) influent organics.
  Because very low  organic concentrations are insufficient to
support  an actively growing biomass, the reactor biofilm slowly
deteriorates (or decays) with time. When appropriate, the biofilm
can be regrown in the reactor through exposure to high organic
concentrations. Specific organic concentrations have been reduced
from the parts per million to the  low /tg/1 range using this reactor
design.

INTRODUCTION
  Three of the most common remediation technologies used to treat
contaminated  groundwater  and  industrial waste waters are  air
stripping, carbon adsorption and biological treatment. Air stripping
is a mass transfer, rather than destruction, technology. Chemicals
dissolved in water are brought into contact with large volumes of
air and the compounds with low water solubilities pass from  the
water phase into the air phase. The contaminants removed from
water are not actually treated; they are just transferred from one
medium to another. Chemicals highly soluble in water (such as
acetone) are removed only to a limited extent. While air stripping
has the advantage of being a relatively inexpensive treatment tech-
nology, increased concern over air pollution and its possible human
health effects is limiting the applicability of air stripping technology.
  Carbon adsorption is a natural process in which molecules of
a liquid or gas are attracted and held at the surface of a solid, such
as carbon. This physical attraction is caused by the surface tension
of the carbon. Organic chemicals have different  affinities  for
carbon, making carbon adsorption more applicable in some cases
than in others. Carbon adsorption is a separation technique,  not
a destruction technology. Spent carbon still retains the organic con-
taminants removed from the contaminated stream and must either
be disposed of as a hazardous waste or regenerated. Carbon systems
have the advantage of being effective as soon as they go on line.
However, costs for replacement carbon and disposal of spent
carbon make this treatment technology relatively expensive. This
is especially true if carbon is used to treat waters at both high flow
rates and high organic concentrations.
  Biological processes (such as activated sludge, trickling filters
and rotating biological contacters) have been used successfully for
many years to treat waters containing high (greater than 50 mg/1)
concentrations of biodegradable influent organics. These treatment
systems foster the  aerobic growth of microorganisms in order to
convert biodegradable contaminant mass into carbon dioxide, water
and additional biomass. Biological treatment of contaminated
waters  containing less than 50 mg/1 was not practical because these
low organic concentrations generally would not support the growth
of additional biomass. Thus, aerobic processes were considered
to have a lower influent threshold concentration of approximately
50 mg/1.
  However, laboratory work by microbiologists and microbial
ecologists showed that organisms can indeed degrade organic com-
pounds to the ;*g/l range1. New  information  about microbial
processes that take place under low nutrient growth conditions also
revealed that biofilm technologies have specific advantages when
applied to the pollution control field. For example, healthy biofilms
grown  at a high specific organic concentration can effectively reduce
the feed organic concentration down to some minimal level, usually
designated as S^  (for minimal substrate level).  If that biofilm is
then switched to a feed concentration less than S^, the target
compound can continue to be effectively scavenged to concentra-
tions far below S^.  Under these conditions, however, the biofilm
does not actively grow  but rather decays with time2  The decay
method of treating water containing low contaminant concentra-
tions was used as  the basis of a research  plan to develop a func-
tional  "decay" biological reactor.
  This  paper describes the theory, development, field testing and
operation of a DETOX "decay" submerged fixed-film bioreactor
designed to treat influent organic concentrations below 50 mg/1.
The non-conventional technology utilized capitalizes on the slow
decay,  rather than growth, of organisms present in a biofilm. A
healthy biofilm is initially grown within the bioreactor using a liquid
recircuJation system and supplemental feed organics. When the bio-
film has sufficiently  matured, the recirculation system is discon-
nected  and the waste stream to be treated (containing low influent
organic concentrations) is fed into the reactor. The decay sub-
merged fixed-film technology is especially applicable to the remedi-
ation of hydrocarbon contaminated groundwaters, such as those
typically found at  sites  containing leaking underground  storage
tanks.

DECAY THEORY
  Decay theory of biodegradation is the culmination of:  (1) bio-
208    CONTAMINATED AQUIFER CONTROL

-------
logical advances made in understanding the growth of micro-
organisms  under low nutrient  conditions, and (2) engineering
advances made in the understanding of fixed-film processes. This
section of the paper will attempt to briefly explain some of the
fundamental biological/engineering processes involved.
  Microorganisms possess a wide variety of metabolic capabilities
and live under many different environmental conditions. While the
almost endless diversity of microbial capabilities may seem con-
fusing to  someone unaccustomed to dealing with biological
processes, all living organisms have the same basic goals. These
goals are (1) to remain alive and (2) to grow and multiply if environ-
mental conditions are favorable.
  The state of being alive requires energy and organic/inorganic
nutrients. Microorganisms use the organic/inorganic compounds
found in their environment as  food to supply themselves with
energy and materials for new biomass. When environmental con-
ditions are favorable, organisms continue to grow and multiply
until conditions change. In general, the growth rate of an organism
is proportional to the concentration of any required factor that
limits growth. Assuming that all nutrients are present in excess,
this growth limitation reverts to the amount of food (or substrate)
available. Using batch culture experiments and non-inhibitory sub-
strates, Monod was among the first to attempt to relate the micro-
bial growth rate to substrate concentration3. He  developed the
empirical relationship:
                      u = U    _ง_                  (1)
  where: u
         Um
         S

         Ks
                Ks + S

specific growth rate (I/time)
maximum specific growth rate (of I/time)
concentration of the    growth-limiting
nutrient in solution (mass/unit volume)
half-velocity constant, that is the nutrient
concentration at one-half the maximum
growth rate (mass/unit volume)
  The growth rate/substrate concentration relationship is shown
in Figure 1. At low substrate concentrations, the growth rate is
also low. As more substrate becomes available (as S increases),
the microbial growth rate increases until some maximum growth
rate (umax) is attained. Later research has shown that two major
types of microorganisms exist:  (1) those with high growth rates
requiring high substrate concentrations (these organisms have high
Ks values) and (2) those with low growth rates that grow best at
low substrate concentrations (as low as  1  mg/1 per day; these
organisms have low KQ values)4-5.
      (U
      o.
      en
           max
           2
                 Limiting Nutrient  Concentration,  S
                            Figure 1
        Monod Growth Kinetics Showing Relationship Between
      Growth Rate (u) and Limiting Nutrient Concentration (S).
                 Ks is the Nutrient Concentration
           At One-Half the Maximum Growth Rate (U   ).
                                                   These findings  have several important  implications for the
                                                 pollution control field. First, treatment of low concentrations of
                                                 organics will most likely be performed by microorganisms that grow
                                                 very slowly. It is important therefore to engineer bioreactors that
                                                 are able to maintain slow growing biomass within the treatment
                                                 system. Fixed-film systems  can maintain sludge ages of 20 to 100
                                                 days, as compared to 4 to 20 days for most activated sludge systems
                                                 with recycle. This sludgeage corresponds to growth rates of 0.3
                                                 day-1 or less, compared to 0.3 to 1.2 day-1 for the aforementioned
                                                 activated sludge systems6. In addition, growth under low nutrient
                                                 conditions seems to favor microbial attachment to surfaces where
                                                 substrate organics may accumulate7. This finding also supports
                                                 the idea that fixed-film processes may be ideal in the treatment
                                                 of waters containing low concentrations of organics.
                                                   Rittmann  and  McCarty8' 9  developed a biofilm model for
                                                 treating low organic concentration solutions such as those found
                                                 in groundwater. The model considered mass transfer of the sub-
                                                 strate through the bulk liquid to the biofilm, diffusion of substrate
                                                 through the biofilm, biological utilization of the substrate, growth
                                                 of the biofilm and decay of the biofilm. They predicted that the
                                                 concentration of substrate needed to keep the biofilm in a steady-
                                                 state (that is, no net gain or loss in the biofilm) condition is given by:
                                                                                                      (2)
                              YK-b

where: Smin = the minimum substrate concentration
              (mass/unit volume)
       Ks   = the Monod half-velocity constant
              (units of mass/unit volume)
       b    = the specific decay rate for the bio-
              film (I/time) and
       Y   = true cell yield (mass of cells
              produced/mass of substrate
              removed)
                                                   Laboratory biofilm reactors using 3 mm glass beads held in 12-cm
                                                 long by 2.5-cm diameter glass columns were used in their tests.
                                                 Once a healthy biofilm had been established, the fixed-film biore-
                                                 actors received a feed solution containing a target chemical (such
                                                 as acetate) as a substrate source. Biological activity occurring in
                                                 the  column as the water passed through the reactor reduced the
                                                 contaminant concentration down to a limiting minimal concentra-
                                                 tion (Smin)  in good agreement with  their steady-state model
                                                 predictions8.
                                                   However, if the bioreactors were operated under nonsteady-state
                                                 conditions in which the feed substrate concentration was below
                                                 the  Smin value,  the biofilm was able to effectively scavenge the
                                                 feed organic to concentrations  much less than Smin  2- 9. A
                                                 nonsteady-state biofilm process using galactose as a substrate was
                                                 able to sustain good (greater than 85%) removal of trace substrate
                                                 concentrations  for  1 year without the need to regenerate the
                                                 biofilm2. An example of acetate removal under both steady and
                                                 nonsteady-state conditions from Rittmann's work is shown in
                                                 Figure 29.
                                                   While the theoretical basis for nonsteady-state biofilm treatment
                                                 of organics was being established, laboratory and pilot-scale work
                                                 on the development of submerged fixed-film bioreactors to treat
                                                 high levels of  organics was also  progressing (10 and  11, for
                                                 example). Submerged fixed-film reactors were small, easily  por-
                                                 table, resistant to shock loads and required a minimal amount of
                                                 operator attention.
                                                   DETOX, Inc. saw that successful development of a decay (or
                                                 nonsteady-state) submerged fixed-film bioreactor would offer new
                                                 opportunities for the pollution control field in treating  ground-
                                                 water and industrial process  waters contaminated with low levels
                                                 of organics. The next sections of this paper present data obtained
                                                                              CONTAMINATED AQUIFER CONTROL     209

-------
during development of the reactor as well as results obtained during
pilot and full-scale field application of the treatment system.
          o.
          o.
          u
          I
          41
8 .
  l
6 .



2 •

0
                       24     6    8    10    12

                      Length  Along Column  (cm)
1.
o.
etate Cone
u
o.a
i
0.6
O.A
0.2
0

V. ,B

              24     6    8   10    12

              Length  Along Column  (cm)

                   Figure 2
     Examples of Acetate Removal Under Both
 Steady-State and Unsteady-State Biofilm Conditions.
Figure 2-A Shows that Under Steady-State Conditions,
Acetate Concentrations can be Reduced from 7.6 mg/1
            To an Smin Concentration
  Of Approximately O.Tmg/I. Figure 2-B Shows (he
  Same Biofilm under Non-Steady State Conditions.
    Influent Acetate Concentrations of 0.76 mg/1
    Can be reduced to less than 0.2 mg/1. S m is
      Denoted  by Dashed Line in Each Graph.
      Adapted from Rittmann and McCarty*.
                                                           for a little over one week (days 50-59), with greater than 98%
                                                           benzene removal. The flow rate was next increased to 7.5 gal/hr
                                                           (hydraulic retention time of 30 min) for 4 weeks (days 60-90) and
                                                           the system continued to achieve excellent (98%) benzene removal
                                                           rates. Finally, the water flow rate was increased to 10 gal/hr (23-min
                                                           hydraulic retention time)  for over 2 weeks (days 91-108). Treat-
                                                           ment efficiencies decreased to approximately 92% during this
                                                           period, presumably due to both the lower hydraulic retention time
                                                           and changes in other operational parameters. Overall, average ben-
                                                           zene removal efficiencies of greater than 96% were attained during
                                                           the 108 day test. This result was achieved in spite of fluctuations
                                                           in organic concentrations, air flow rates and water flow rates.
                                                                                                                     M   100   1U
                                                                                               Figure 3
                                                                        Benzene Removal During Laboratory Bioreactor Development.
                                                                              Key: Influent Benzene (•);  Effluent Benzene (o).
                                                                     After this first round of benzene experiments was completed,
                                                                   the biofilm within the reactor was regrov. n in anticipation of the
                                                                   second round of tests investigating the biodegradability of low con-
                                                                   centrations of MEK. Flow to the reactor was reduced to 2.5 gal/hr
                                                                   for 4 weeks and benzene was again fed into the reactor.
                                                                     The test reactor was then operated for 9 weeks at MEK concen-
                                                                   trations of 2-10 mg/1 and flow rates of  2.5-4.5 gal/hr  (corres-
                                                                   ponding  to  hydraulic   retention times  of 90 and 45  min,
                                                                   respectively). Removal efficiencies were consistently greater than
                                                                   98% throughout the test period (Fig. 4).
LABORATORY DEVELOPMENT OF DECAY REACTOR
  A laboratory reactor using stacked packing material as a support
for biofilm growth was designed. The test column was 0.3 ft in
diameter and 6 ft tall. Water pumped to the reactor was initially
saturated with oxygen through the use of air stones placed in a
holding tank. After it had been pumped from the holding tank,
a concentrated feed stock of the organic substrate was then metered
into the aerated water. Water  flowed down through the column
while air introduced into the bottom of the reactor passed up and
out.
  Approximately 2 months were required to initially establish a
benzene degrading biofilm within the stacked-pack reactor. Over
the next 200 days, the reactor was used  to test  the effects that
hydraulic retention time, air flow rate and other parameters had
on  the biodegradation of low (less than 10 mg/1) concentrations
of benzene and methyl ethyl ketone  (MEK).
  The overall results obtained using benzene as the target substrate
are shown in Figure 3. Over the initial 108 day test period, the flow
rate of water to the reactor was increased from 2.5 to 10 gal/hr
while maintaining a relatively constant benzene concentration.
From days 0 to 31, the reactor received a flow of 2.5 gal/hr, corres-
ponding to a hydraulic retention time of 90 min. Different air flow
rates  were tested during this period and overall benzene  removal
rates greater than 96% were attained. The reactor was next operated
(days 32-49) at a flow of 3.5 gal/hr, giving a hydraulic retention
time of 64 min. As before, different air flow rates were tested and
removal efficiencies of greater than 97%  were observed. A flow
rate of 5 gal/hr (hydraulic retention time of 45  min)  was tested
                                                                                      Days of Operation


                                                                                       Figure 4
                                                                 MEK Removal During Laboratory Bioreactor Development.
                                                                        Key: Influent MEK (•); Effluent MEK (ฐ}.
                                                              Data obtained during laboratory bioreactor development con-
                                                            firmed that biological decay processes could effectively treat low
                                                            concentrations of environmentally significant chemicals. Reaction
                                                            rates were fast enough to allow treatment hydraulic retention times
210     CONTAMINATED^ AQUIFER CONTROL

-------
as low as 30 min. In addition, certain proprietary methods were
developed during this period to extend the usual life of the biofilm.
Thus, the theory and technology associated with biological decay
reactors appeared to be viable and awaited field testing under real
world conditions.

FIELD PILOT TESTING OF REACTOR
  An opportunity arose to pilot-test a prototype decay reactor in
the field at an industrial site in New Jersey. Water discharged from
the facility periodically exceeded effluent standards set for benzene
(200 /tg/1), BOD and TSS. It was hoped that the bioreactor could
consistently reduce benzene levels to below 200 /ig/1 and also reduce
BOD  concentrations.
  Water from the facility consisted of both contaminated ground-
water and various plant process waters, including well filter back-
wash and cooling tower wash waters. The actual total organic and
inorganic composition of the water was not available. It was known
that benzene concentrations typically varied from 240 to 1,400 /tg/1.
A diagram of the prototype DETOX reactor used  at the site is
shown in Figure 5. The pilot study was conducted over a 75-day
period (December 1986 to February 1987) and tested  the effects
of three different hydraulic retention times and two different mixing
schemes within the reactor. 'Benzene  data from the study  are
presented in Figure 6.
                                        Air Outlet
     Influent
       Water  ฃ
        Compressed
          Air Line

        Air
                                               Effluent
                                                Water
                        T""   /
                          \ Underdrain /
                        Collection System
                            Figure 5
   Diagram of Prototype DETOX Decay Submerged Fixed-Film Reactor
                                                           Overall, benzene removal efficiencies of greater than 89% were
                                                         attained. For the first 37 days of the test, the reactor was operated
                                                         in a plug-flow mode with an influent flow of 4-5 gal/hr, giving
                                                         a hydraulic retention time of approximately 90 min. From days
                                                         38 to 61, the reactor operated in a plug-flow mode at 2-2.5 gal/hr,
                                                         or a hydraulic retention time of approximately 180 min. Lastly,
                                                         aeration to the reactor was increased and the system was operated
                                                         for 14 days in a completely mixed fashion at a flow rate of 1 gal/hr
                                                         (hydraulic retention time of 360 min). Benzene removals through-
                                                         out the test were consistently good, but BOD, COD and TSS values
                                                         varied widely and did not meet discharge criteria (data not shown).
                                                           Upon further evaluation of the test results, it appeared "that the
                                                         plant water was inhibitory to suspended growth microorganisms
                                                         and that the fixed-film system actually reduced at least part of the
                                                         inhibition. This  hypothesis was supported by routinely finding
                                                         higher BOD concentrations in water tested after treatment (data
                                                         not shown). While the prototype system could not meet the dual
                                                         demands of the client (both benzene and BOD reductions), it did
                                                         'demonstrate that the decay biological reactor could effectively treat
                                                         low concentrations of environmentally significant chemicals under
                                                         real-world conditions.

                                                         FULL SCALE TREATMENT OF HYDROCARBON
                                                         CONTAMINATED GROUNDWATER
                                                           Improvements were made to  the design of the decay biofilm
                                                         reactor following the field pilot  testing. In January 1987, one of
                                                         the first DETOX L-Series decay submerged fixed-film reactors was
                                                         used to remediate a site in California (south of San Francisco) in
                                                         which groundwater became contaminated with gasoline as the result
                                                         of a leaking underground storage tank. The client believed that
                                                         total organic concentrations would be  in the 25 mg/1 range with
                                                         flow rates less than 300 gal/hr. The proposed treatment system
                                                         consisted of a groundwater recovery well, above ground oil-water
                                                         separator, bioreactor, sump, roughing filter and activated carbon
                                                         polishing filter. The activated carbon filter was needed to meet strin-
                                                         gent California water discharge criteria, such as effluent benzene
                                                         concentrations of 0.7 /tg/1 or less.
                                                           The DETOX L-6 reactor was installed and started up in early
                                                         1987. When the treatment system was put on line in March 1987,
                                                         influent total hydrocarbon concentrations were in the 250-270 mg/1
                                                         range, far above the maximum design concentration of an L-Series
                                                         reactor (25-50 mg/1). However,  the bioreactor was able to adapt
                                                         to the higher organic concentrations by functioning as a growth
                                                         reactor, not as the anticipated decay reactor. Since March, the in-
                                                         fluent total hydrocarbon concentration to the system has steadily
                                                         declined and in June typical concentration values were approxi-
                                                         mately 50 mg/1. Throughout the 100 days of operation thus far
    o
    c
    o
    CJ
    0)
    c
    0)
    M
    C
    a
    PQ
        1200-.
         800-
400.
            0          40        60         80

                          Days of Operation


                           Figure 6
         Benzene Removal Using Prototype Decay Reactor.
         Key: Influent Benzene (•); Effluent Benzene (ฐ).
                                                     100
                                                                    o
                                                                    o
                                                                    u
                                                           o
                                                           H
                                                                        300
                                                                        200  .
                                                                100
                                                                                                      60
                                                                                 Days  of Operation
                                                                                                              80
                                                                                                                       100
                                                                                    Figure 7
                                                              Total Hydrocarbon Removal Using DETOX L-6 Bioreactor to
                                                                     Treat Gasoline Contaminated Groundwater.
                                                          Key: Influent Total Hydrocarbons (• ); Effluent Total Hydrocarbons (o;.

                                                                     CONTAMINATED AQUIFER CONTROL    211

-------
(March 18,1987 to June 20,1987), the system has removed greater
than 90% of the total hydrocarbons present. During this period,
benzene concentrations were reduced by more than 93%, toluene
concentrations by more than 96% and xylene concentrations by
more than 91%. Operating data for total hydrocarbon and benzene
removals are shown in Figures 7 and 8, respectively.
        14
    I.
    a.
                     20
   40      60
Days of Operation
                                                        100
                            Figure 8
          Benzene Removal Using DETOX: L-6 Bioreactor to
             Treat Gasoline Contaminated Groundwater.
           Key: Influent Benzene (•); Effluent Benzene (ฐ).

   It is important to realize that accurate design data (such as in-
fluent organic concentration) frequently are unavailable when treat-
ment equipment is sized and sold. Thus, a treatment system that
has flexibility in terms of handling influent concentrations and flow
rates may not only be desirable but required. As groundwater
pumping continues at this site, influent organic concentrations are
expected to continue to decline. As they do, the biological processes
within the reactor will shift from growth to decay mode. This will
allow effective treatment to continue at the site using the same treat-
ment equipment. However, since growth mode reactors cannot
typically reduce specific organics to the low /tg/1 range, larger
amounts of activated carbon will be used for polishing as long as
the reactor is operated in  the growth mode.
   Decay  reactors  use  aerobic  biofilm  processes and must  be
supplied with minimal amounts of air during operation. Because
of strict air emission requirements in California, state regulatory
personnel were concerned that some of the volatile gasoline ground-
water contaminants were being removed by air stripping rather than
by biological processes. Off gases emanating from an air vent in
the top of the covered bioreactor were sampled in triplicate on June
3, 1987 for total hydrocarbons, carbon dioxide and oxygen using
Source Test Method 1-100 of the California Air Resources Board.
Carbon dioxide from the reactor was continuously monitored for
three 30-min periods using a Horiba Model PIR-2000 NDIR car-
bon dioxide analyzer. Total hydrocarbons were quantitated simi-
larly using a Beckman Model 400 Hydrocarbon analyzer and
oxygen concentrations were measured with an Infrared Industries
Model 2200 Oxygen analyzer. Estimated air flow  rates from the
system were  1 ftVmin.
   Total air hydrocarbon concentrations (as C-l) were determined
to be  300,  371  and 423  mg/1. Methane  concentrations  were
measured at  2, 2 and 2 mg/1. Total non-methane hydrocarbons
(as C-l) were measured at 298, 369 and 421 mg/1. Oxygen concen-
trations were close to normal (21.6, 20.2 and 20.6%) and carbon
dioxide concentrations  varied  (0.32, 0.60 and  0.95%).  Total
hydrocarbons (as  C-l) released to the  atmosphere averaged
0.000678 Ib/hr. Total non-methane hydrocarbons (as C-l) released
to the atmosphere averaged 0.000674 Ib/hr.
   Unfortunately, a mix-up with the water sampling crew occurred
and water samples were not taken while the air monitoring was
being  done.  This mistake prevented a  total contaminant  mass
balance analysis from being performed on the  treatment system.
Water samples taken the next day (June 4, 1987), however, can
 provide at least an estimate as to the treatment efficiency of the
 system. Assuming that contaminant concentrations were approxi-
 mately the same over the 48-hr period, 0.44 Ib of total hydrocarbons
 were calculated to have passed through  the system per day at a
 flow rate of 1.94 gal/min and the total hydrocarbon concentra-
 tion in the water was 19 mg/1. (This value for hydrocarbon con-
 centration was the  lowest observed to date). Using the average total
 hydrocarbon (as C-l) value released to the atmosphere of 0.000678
 Ib/hr, approximately 0.0162 Ib of hydrocarbons are released into
 the air per day from the bioreactor. This amount corresponds to
 an air stripping rate of approximately 3.68% of hydrocarbons
 entering the tower. The minimal rates of air flow to the biotreat-
 ment system help to ensure that readily biodegradable volatile
 organics are biodegraded, not air stripped.
   Concern over the limited air discharge from the site does not
 appear to be warranted at this time, as  the performance of the
 bioreactor was evaluated and approved by the appropriate Cali-
 fornia regulatory personnel. Further treatment of the off-gases was
 not deemed necessary.
CONCLUSION
  Decay  fixed-film reactors  can be designed and operated to
biodegrade low (less than 50 mg/l) influent organic concentrations.
These low starting concentrations  are less than the treatment
thresholds associated with typical biological growth mode reactors.
  Expanding on concepts developed for nonsteady-state bio films,
a healthy biofilm is initially grown within the reactor using a water
recirculation system and high influent organic concentrations. Once
the biofilm is established, the reactor can be switched to a feed
containing low  organic concentrations and the biomass can con-
tinue to reduce these compounds to the low jig/1 range. With time,
this  biofilm slowly decays, and  eventually must  be regrown to
continue effective  treatment.
  DETOX,  Inc. has developed, tested and installed submerged
fixed-film reactors  utilizing the biological decay concept. Benzene,
toluene, xylene, methyl ethyl ketone and petroleum constituents
are especially amenable to remediation using this technique.  The
ability of the decay reactor to treat low influent organic concen-
trations,  such as those typically  found in groundwater or dilute
industrial process waters, makes it a valuable tool for use by the
pollution  control field.

REFERENCES
  1. Alexander, M.,  "Biodegradation of Organic  Chemicals," Envir.
    Sci.and Tech.. 18. 1985, 106-111.
  2. Rittmann, B.E. and Brunner, C.W., "The Nonsteady-State, Biofilm
    Process for Advanced Organics Removal," JWPCF, 56, 1984, 874-880.
  3. Monod, J., "The Growth of Bacterial Cultures," Ann. Rev. Micro..
    3. 1949, 371-394.
  4. Hirsh, P., "Life Under Conditions of Low Nutrient Concentrations",
    in Strategies of Microbial L(fe in Extreme Environments, M. Shflo,
    Ed., Verlag Chemie,  Berlin,  1979. pp. 357-372.
  5. Harder, W. and Dijkhuizen, L., "Physiological Responses to Nutrient
    Limitation," Ann. Rev. Micro., 37, 1983, 1-23.
  6. Kobayashi, H. and Rittmann, B.E., "Microbial Removal of Hazardous
    Organic Compounds." Envir. Sci.  Tech.. 16.  1982. 170A-183A.
  7. Kjelleberg, S., "Effects of Interfaces on Survival Mechanisms of Copfo-
    trophic Bacteria in Low-Nutrient Habitats," in Current Perspectives
    in Microbial Ecology, M. J. Klug and C. A. Reddy, Eds., American
    Society for Microbiology, Washington, D.C., 151-159, 1984.
  8. Rittmann, B. E. and McCarty, P. L., "Evaluation of Steady, State-
    Biofilm Kinetics," Biotech. Bioeng., 22,  1980, 2359-2374.
  9. Rittmann, B. E. and McCarty, P. L., "Substrate Flux Into Biofilms
    of Any Thickness," J. Environ. Eng. 107,  1981, 831-849.
 10. Rustin, B., "Wastewater Treatment with Aerated Submerged Biologi-
    cal Filters," JWPCF56, 1984, 424-431.
 11. Lee, K. M. and  Stensel. H. D.,  "Aeration and Substrate Utilization
    in a  Sparged Packed-Bed Biofilm Reactor," JWPCF,  58,  1986,
    1066-1072.
212     CONTAMINATED AQUIFER CONTROL

-------
          Evaluation of Groundwater Remediation Techniques   for
                 Fractured  Bedrock Using  Aquifer  Response Tests
                                                Kenneth A. Wallace
                                        Environmental  Protection Agency
                                                    Region VIII
                                                 Helena, Montana
                                                Paul  L. Karmazinski
                                                 NUS  Corporation
                                       Waste Management Services Group
                                                Clearwater, Florida
                                                 Douglas J. Yeskis
                                     U.S. Environmental  Protection  Agency
                                                      Region V
                                                  Chicago,  Illinois
ABSTRACT
  Aquifer response (pumping) tests of two drinking water aquifers
underlying the Byron/Johnson Salvage Yard hazardous waste site
were conducted to determine hydrogeologic conditions and evaluate
the feasibility of alternative groundwater remediation techniques.
The pumping tests and down hole geophysical logging were
necessary to define the complex hydrogeologic conditions at the site.
  While hydrogeological interpretation was the primary reason for
conducting pumping tests, they also provided an appropriate stage
to evaluate potential groundwater treatment techniques. Hydraulic
data derived from the tests are being used to determine the via-
bility of pump and treat aquifer purge systems. A pilot scale treat-
ment system was installed on the pumping system to determine the
effectiveness of various contaminant removal methods and to
permit surface disposal of purge water.

INTRODUCTION
  Possibly the greatest threat to human health and the environ-
ment  from  hazardous wastes is created  by the movement of
pollutants into the water resources. Once contaminants are released
into an aqueous environment, toxicity reduction and even contain-
ment rapidly become less achievable. Restoration of contaminated
bedrock aquifers is among  the most difficult problems facing
environmental remediators.  To make viable engineering evalua-
tions of cleanup alternatives and to provide accurate cost analyses
of those alternatives, detailed geologic and hydrogeologic infor-
mation must be compiled and then interpreted. Pumping tests and
treatability studies are critical components of the planning and
evaluation process undertaken prior to design work for aquifer
remediation.

SITE  DESCRIPTION
  The Byron/Johnson Salvage Yard Superfund site is located
approximately 12 miles southwest of the city of Rockford in north
central Illinois. The 20-acre  site is situated within the Woodland
Creek drainage basin and consists of mostly uplands dissected by
several small north and northeast trending ravines. The ravines feed
into Woodland Creek, a tributary to the  Rock River.
  The site operated during the 1960s and early 1970s as a salvage
yard and unpermitted landfill. The Illinois Environmental Pro-
tection Agency (IEPA) began investigating the site in 1970, but
illegal dumping activity occurred until 1977,and reports of con-
taminant discharges  continued until  the  early  1980s.  The
Byron/Johnson Salvage Yard was placed on the NPL in 1982.
  Soils found on the  Salvage Yard were contaminated by lead,
nickel, zinc,arsenic,  cyanide and halogenated organics.  Con-.
taminants found in surface and buried drums included lead, arsenic,
cyanide, halogenated  organics and low level polychlorinated
biphenyls. Groundwater at the site and hydraulically downgra-
dientis  contaminated with heavy metals,  cyanide and volatile
organics. Sampling of private  wells downgradient from the site
revealed high concentrations (up to 710 /tg/1) of trichloroethylene
(TCE) in some locations. Residents in the immediate vicinity of
the Salvage Yard were placed on a bottled water  program; they
later received carbon filters designed to remove volatile organics
from all water entering the household. Further sampling of private
wells in a subdivision approximately 1.5  miles northwest  and
hydraulically downgradient of  the site indicated that some water
supplies also were being affected by the  TCE contamination,
although at much lower concentrations. A supplemental RI/FS was
initiated in June  1985 to evaluate the groundwater  contamination
problems emanating from the  Salvage Yard.

SITE GEOLOGY
  The Salvage Yard is covered  by a relatively thin layer (< 10 ft)
of soil sand unconsolidated Pleistocene glacial and/or fluvio-glacial
deposits which overlay the two bedrock aquifers of primary interest.
These aquifers provide drinking water to residents living near and
downgradient of the Salvage Yard site. The unconfined upper
aquifer is composed of fractured dolomites of the Galena-Platteville
Group. The lower, semi-confined aquifer consists of homogeneous
St. Peter sandstones. The two aquifers are separated by a relatively
thin bed of shales, the Harmony Hill member of the Glenwood
Formation.1  The Harmony Hill  acts,  to some  extent,  as an
aquitard between the upper and lower aquifers. Other aquifers are
found beneath the St. Peter, but no downgradient wells are screened
in the deeper formations to document whether contaminants from
the Salvage Yard are present.
  Because contaminant levels decrease with depth and the discharge
point for the upper aquifers is the Rock River, contamination from
the Salvage Yard  is not expected in deeper formations.  Ad-
ditionally, domestic use downgradient from the Salvage Yard is
limited to  the Galena-Platteville and St. Peter aquifers.1
  The Salvage Yard is situated on a hydrogeologic divide. Ground-
water flows: northwest toward a subdivision; west toward a sparsely
populated rural area; and north  by northeast toward the Woodland
Creek drainage, an inferred primary fracture trace and structural
trend.2 All flow ultimately discharges into the Rock River.
  The complex subsurface conditions at the Byron site warranted
a hydrogeologic investigation.  This was conducted by the Illinois
District of the U.S. Geological Survey (USGS) in conjunction with
the U.S. EPA, Region V, Emergency and Remedial Response

          CONTAMINATED AQUIFER CONTROL    213

-------
(Marclch.3 The ultimate purpose of the investigation was to guide
thaie U.S. EPA in the Superfund remedy selection process, but
b certain critical hydrogeologic characteristics are needed to make
 definitive evaluations of groundwater cleanup technologies. The
 goals of the hydrogeologic evaluation were to determine: (1) the
 extent of hydraulic connection between the two aquifers, (2) the
 preferred fracture-flow directions and  contaminant migration
 routes, (3) the storativity and specific yield of the respective aquifers
 and (4) the transmissivities of the two aquifers.

  PUMP TEST DESIGN
    Borehole geophysical testing of selected wells on and near the
  Salvage Yard was performed to characterize lithologies. Caliper
  and acoustic televiewer logs of uncased wells indicate the presence
  of large horizontal fractures in the  Galena-Platteville dolomite
  units. Porosities of the lithologies, as determined by laboratory
  testing of rock core samples, ranged from 8 to 15% in the dolomites
  and 14 to 17<% in the St. Peter Sandstones.4
    Step-drawdown tests were conducted  in each aquifer in  order
  to design the aquifer response tests. Each aquifer was pumped for
  set periods of time at increasing rates,  and drawdown was observed
  for each rate through time. The purpose of the step-drawdown tests
  was to determine the maximum sustained pumping rate for each
  well.  Maximum stress is  therefore applied to  the aquifer,  con-
  strained by the physical limitations of the aquifer, the depth and
  diameter of the pumping well, and the  pump  used.
    The results of the step-drawdown  tests defined the conditions
  and duration of the pumping tests for each aquifer. Maximum sus-
  tainable pumping rate determined for  each aquifer was 20 gal/min.
  It was hoped that the upper aquifers' fractured nature would create
  much higher flow rates. However, the fractures within the aquifer
  may have undergone significant porosity reduction because of solu-
  tion precipitation and clay in filling. Alternatively, the well may
  not have intercepted major  water bearing fractures.
    The USGS also used the data from  observation wells monitored
  during the step-drawdown tests to determine the duration of the
  pumping tests. The lower St. Peter and upper Galena-Platteville
  aquifers were to be pumped at 20 gal/min for 2 and 4 days, respec-
  tively, to ensure sufficient drawdown in observation wells. Water
  level recovery  in the wells also would monitored for  an  equal
  amount of time.

  TREATMENT SYSTEM OBJECTIVES
    The U.S. EPA Environmental   Response  Test (ERT) was
  requested by Region V to address the problem of pumping well
  effluent disposal.  ERT tasked  the  Environmental  Emergency
  Response Unit (EERU) to assist in the project. Three methods for
  disposing  pumping well  effluent were  considered: local  sewer
  systems, trucking to a licensed disposal site or diversion to a natural
  drain. (A fourth method, purge water storage in on-site tanks, was
  dismissed earlier because of initial estimations that over 576,000
  gal of pump well effluent  would need to be contained). There are
  no local sewer systems available, and the costs and logistics involved
  with a truck and disposal system were  prohibitive. Effluent disposal
  to a natural surface drainage was the only workable  alternative.
  Purge water from the pumping tests could not be discharged in
  the vicinity of the pumping or observation wells because of the
  potential creation of artificial recharge conditions. In order  to
  release the effluent away from the zone of pumping influence off-
  site discharge was required. Pre-treatment prior to off-site discharge
  was necessary because of  the potentially high contaminant levels
  that could be present in the  purge water.
    The design and implementation of a transportable water treat-
  ment  system for the Byron pump tests had a two fold purpose.
  Treatment of pumping well effluent prior to discharge was required
  to obtain an IEPA Water Pollution  Control Permit.  The second
  objective was to implement a pilot scale system that would test
  techniques potentially applicable to  a full-scale pump and treat
  system.
TREATMENT SYSTEM DESIGN
  Preliminary design of the treatment system was based on a sum-
mary of highest contaminant concentrations found in monitoring
wells in both aquifers. The indicator chemicals used in the design
were volatile organic compounds trichloroethylene (760 pgA), tetra-
chloroethylene (100 ng/l), and vinyl chloride (200 pg/I) with a com-
bined concentration of approximately 1060 /ig/1; metals, arsenic
(102 /ig/1) and nickel (150 Mg/1). with a maximum combined con-
centration of  252 /ig/1; and cyanide at a maximum concentration
of 250 Mg/1.
  Pumping and sampling results from the step tests refined the
design parameters. A maximum design pumping rate of 25 gal/min
was selected,  with a maximum groundwater treatment volume of
216,000 gal for both tests. Time  series sampling was performed
on the pumping wells during the step tests. These samples indicated
relatively  uniform concentrations of indicator  compounds
throughout the tests (Table 1). Concentrations of all parameters
from the time series sampling of the Galena-Platteville aquifer were
below maximum site values, with the exception of vinyl chloride.
Concentrations of all parameters from the time series sampling of
the St. Peter aquifer were well below the highest values for the site.
                          Ttbk 1
                 Results of Sampling Conducted
                 Daring Step-Drawdown Testing
 HUlllOT Lปปl
               H-l-l tl-1-1 n-l-l O-H It-H
                                          P-ซ-l  ป-ป•!
                                                      B+4
                                ซ.*!

                                9.BM
               •lu I
  Although contaminant concentrations from the step-tests were
relatively low compared to the highest historical values, the treat-
ment system design was based on maximum contaminant concen-
trations for two reasons. First, sampling techniques used during
the step tests may have increased contaminant volatilization and
therefore  produced  uncharacteristically  low  concentrations.
Second, the potential for drawing in higher contaminant concen-
trations during extended pumping justified a conservative design
approach.

TREATMENT SYSTEM COMPONENTS
  A design criterion for the system was the rapid assembly of
readily available components using standard fittings. Quick connect
fittings were used to link all  components except for dual routing
to permit bag filter changes. The system used single phase 220 v
as a power source. Shop assembly  of the components required
21 man days. Field assembly took a two man crew 4 hours.
  Granular activated carbon was chosen to remove volatile organic
compounds (VOCs) other than vinyl chloride, the only VOC iden-
tified that  is  not readily adsorbable. Vinyl chloride is extremely
amenable to  aeration and for that reason stripping towers were
included in the system. Mixed bed ion exchange resins were chosen
as polishing agents for the metals and cyanide. All system compo-
nents were designed to treat higher than predicted contaminant
concentrations.
  214     CONTAMINATED AQUIFER CONTROL

-------
Granular Activated Carbon
  Two Tigg C-50 units, each containing approximately 950 Ib of
carbon providing a 14 min retention time per unit, were used.
Pumping well effluent was transferred from Tank 1 with a float
activated 1 hp centrifugal pump through two Rosedale Bag Filter
systems (Figs. 1 and 2). Filtering before the carbon units prevented
sediments from plugging the carbon beds. The bag filters were
placed in series,  using 3  then 1 micron filter bags, filtering well
into the clay fraction. After filtration, the effluent passed through
the carbon  units into Tank 2.
                          Figure 1
          Byron Salvage Transportable Treatment System
                        Flow Diagram
    , H now PROM ruynw
                                                OUT FLOW TO
                                             ACTIVATED CAFIIOK UMTป
       - NORMAL PLOW

       ป ALTERNATE PLOW TO CHANQC IAO FILTERS
                            Figure 2
                  Schematic Dual Bag Filter System
 Mixed Bed Ion Exchange System
   Two options were considered for metals and cyanide removal,
 activated alumina and mixed bed ion exchange resins. Bench scale
 laboratory tests were designed to determine the more effective treat-
 ment system using water collected during the step drawdown tests.
 Columns were packed with sand (bag filter substitute) and granular
 activated  carbon. The flow was then split through the activated
 alumina and the resins. The system was run twice, treating effluent
 from each well separately at retention times equal to the full scale
 system.
   Concentrations before treatment were very close to detection
 limits  for both potential systems;  consequently, we  could not
 determine which was the more effective treatment system. Amber-
 lite mixed bed ion exchange resins were chosen because a complete
 system was available within a short time frame. Ten cubic feet of
 resins were  used, doubling  the  recommended  ratio  of
 5  gal/min/ft3 resin. Effluent was transferred from Tank 2 with
 a  float activated 1 hp centrifugal pump through a l-/t Rosedale
 bag filter system to remove carbon fines, through the ion exchange
 unit to Tank 3.
Packed Column Air Stripping
  The packed column air  stripper was added to  remove vinyl
chloride and unremoved VOCs that might pass through the carbon
treatment system. A float activated 1 hp centrifugal pump trans-
ferred water from Tank 3 to the stripping towers (Figure 1). Flow
was split prior to reaching  the stripping towers. Influent passed
through two 10 foot packed columns in parallel flow and co-treated
with an approximate 14:1 air/water ratio by design. Effluent was
gathered in sumps at the base of each tower and transferred to
Tank 4 by a centrifugal and a sump pump. Final discharge from
the system was by a float activated 5 hp centrifugal pump trans-
ferring the treated water from Tank 4 to a natural drainage channel
approximately 1,200 ft away.

PUMPING TEST RESULTS
Aquifers Response
  The St. Peter aquifer was pumped for approximately 48 hrs at
a rate of 20 gal/min. During the deep well pumping, two observa-
tion wells  screened in the  St. Peter and  five observation wells
screened in the upper Galena-Platteville plus the shallow pumping
well, were monitored (Fig 3). The same wells, plus the pumping
well, were monitored for two days after pumping ceased.
  Pumping on the shallow well was originally scheduled to continue
for 4 days  at a rate of 20 gal/min, but testing was halted after
approximately 52 hr because of dewatering within the pumping
well. Three wells screened in the St. Peter,  including the deep
pumping well, and nine wells screened in the Galena-Platteville were
monitored  during pumping of the shallow aquifer (Fig.  3). The
same wells, plus the pumping well, were monitored for at least
48 hr during the recovery period. Some wells were monitored for
up to 108  hours after pumping ceased.
  Water levels were monitored using pressure transducers rated
from 5 to  15 lb//*2,  and recorded on single and  multiple data
tracking systems. Manually controlled voltage regulators were used
as part of the system. Water levels were recorded on a logarithmic
scale once pumping or recovery had begun. Backup copies of data
                                                                       6W-I60
         GW-42
                      X  ORIENTATION  OF MAXIMUM
                         DRAWDOWN DURING SHALLOW
                         PUMP TEST (line A  to  B)

                      • WELLS  MONITORED  DURING
                          SHALLOW PUMP  TEST

                      D WELLS  MONITORED  DURING DEEP
                                 PUMP  TEST

                      A OFF SITE BACKGROUND WELLS
                      O SHALLOW PUMP  WELL
                      • DEEP PUMP  WELL
                 B
                  • .
                    SALVAGE
                         YARD
                             JBL
                               0'—"™200
                               SCALE (fซซt)
                                                    A  B
                            Figure 3

             CONTAMINATED AQUIFER CONTROL     215

-------
were recorded manually.  Occasionally, water levels were taken
manually to insure maintenance of accuracy  in  the  pressure
transducers.
  Two off-site wells were monitored on an hourly basis with a
Steven's float recorder before, during and after the pumping tests
(Fig 3). One of these wells (GW-160) was open to both the St. Peter
and Galena-Platteville aquifers while the other well (OW-42) was
open only to the Galena-Platteville. Barometric pressures were
taken  from the NOAA weather station located  at the Rockford
airport approximately 12 miles northeast of the site.

Treatment System Performance
  Well effluent was screened periodically during the pumping tests
to ensure discharge water quality and measure treatment efficien-
cies of various components of the system. Screening was conducted
with a Photovac portable Gas Chromatograph calibrated for
various volatile  organic compounds. Additional samples  were
collected from both aquifers and shipped to a U.S. EPA Contract
Laboratory to undergo analyses for volatile organic compounds,
metals and cyanide. Samples were collected  in series to evaluate
contaminant concentrations in the raw pump water and after each
treatment stage, through time.  The results of the field screening
program for volatile organics within the upper aquifer are presented
in Table 2. (Most samples from the  lower aquifer contained  no
detectable concentration  of volatile  organics). Data  for  all
parameters analyzed from the CLP sampling were not available
at manuscript submittal.

                          Table 2
       Results of Field Screening  for Select Volatile  Organic
       Compounds In The Shallow  Aquifer Pumping Well*
mi-i
0*1-2
DM-J
DM1 -4

0*2-1
D*2-l-Ouo
0*2-2
0*2-3
0*2-4
OV2-S

OWJ-1
0*3-2
0*3-3
DM3-4
OK3-5

0*4-1
0*4-2
0*4-3
0*4-4
0*4-5

0*6-1
DWS-1-Oup
DUS-2
DNS- 3
0*S-4
0*5-5

0*6-1
0*6-2
OU6-2-Oup
DW6-3
0*6-4
0*6-5
Date Vinyl
Tie* Chloride
6797BJ 	
2000 hrs 117



6/1
0701





6/1
130




6/1
210




6/1
070C





6/1
190C




J
1
2
/87
hrs 172
183
-
.
1
.
/87
hrs IK
2
1
-
2
/87
nn 112

-
1
.
/87
hn 219
279
.



/87
hrs 104
.
I



                           Bengene  TCฃ  Toluene

                                   6     352
                                         4
                                   1      1
                                        269
                                        2M
                                       286
                                         3
                                         1
                                       211
                                         2
                                       194
                                       179
                                       1S7
                                         I
 551
 525
1203
  2
                                                    Elnyl-
                                                    Benlene
 96
 89
241
 2
 384
 555
485
  1
 69
 95
   Description: Six sett of saoplet Mere collected.  Samples are lilted fn sequence.
   the luutber foil wing the dash Identifies the sanple locitlon, es fnlloot:
   1 - Untreated Influent; 2 • After Carbon Filters; 3 • After Ion Exchange Colimn;
   4 ซ After Air Strippers; end 5 - Sample Taken at Discharge Point In woodland
   Creek.

   •All tairples parts per billion.
  Approximately 57,600 gal of groundwater from the St. Peter
aquifer were treated and discharged. Formation water from the
St.  Peter aquifer had little detectable contamination.  The only
measurable  organics were  encountered after the  ion  exchange
process. Data from previous samplings of monitoring wells on and
downgradient from the Byron Salvage Yard  indicate that forma-
tion water in the St. Peter Sandstone is contaminated, although
at much lower concentrations than the upper aquifer. The lack of
contamination in the deep pump well may be explained by its
position on the site, situated in a relatively upgradient position.
Contaminants migrating vertically from  the unsaturated zone,
through the upper aquifer, through a semi-impermeable boundary
and finally into the St. Peter would be controlled by a horizontal
flow gradient. The proximity of the pumping well to a source of
contamination, and the hydraulic barrier of the aquitard, may not
allow the vertical flow component to overcome the horizontal flow
component. Contaminants may not reach the St. Peter aquifer until
they have already migrated downgradient of the lower pump well.
  Approximately 62,400 gal from the upper  Galena-Platteville
aquifer were treated and discharged.  Samples from the upper
aquifer exhibited relatively high concentrations of volatile organics,
including additional  compounds  not  identified from previous
monitoring well samplings (Table 2). Treatment efficiencies were
greater than 99% for total volatile organics throughout the dura-
tion of the upper aquifer pumping test. Design safety factors com-
pensated for the higher total concentrations and additional com-
pounds encountered. The air strippers  were unnecessary for
removal  of vinyl chloride in  these treatability tests. However,
removal of vinyl chloride using activated  granular carbon is not
cost effective,  and air strippers probably still would be considered
the most viable treatment alternative for this phase of volatile
organic contamination.

HYDROGEOLOGIC EVALUATION
  The  extent of hydraulic connection between the two aquifers was
determined using the Hantush-Jacob solutions to determine the
leakage coefficient.  The data best matched semi-confining curves,
which  then were used in the analysis. Leakage values obtained
ranged from  3.1 x  10"J to 1.6 x 1O4 ft/day/ft.
  Drawdowns observed during the Galena-Platteville pump test
indicate  a groundwater  flow  trending  southeast to northwest,
roughly parallel to primary fracture directions.2 This orientation
can be observed in Fig 3. Drawdowns in wells along the trend A
to B are uniformly greater when compared to wells  the same
distance  from the pumping well but in other orientations.
  Transmissiviries and storage coefficients for each aquifer were
determined from time-drawdown curves developed for each well
and analyzed using Theis curve  matching techniques,  double
porosity analysis and Jacob straight-line methods. The transmis-
sivity and storativity or specific yield for each aquifer is presented
in Table 3. The numbers represent the range of values determined
by each analytical technique. Seven wells were used for the Galena-
Platteville determinations while two wells were used for the St. Peter
aquifer.
                                          Table 3
                            Transmisslvitles and Storage Coefficients
Kin?* of
•"<ซ "•""-
T 1, 900 to ll.OOO"

I 11. 000 to 14,000
11. ปซ>r
S ).i i 10'* to 1.0 I 10*>
..„, 	 /ซ,/„ e.t.r.1
Th*U Nซtch
S.9S4 to 1 Sf, 000

t..)QO to 11,000
6.0 ป 10-* tO 1,8 I lQ->
ttrt by:
Ml. ^^
2,000 to 4,m

tot (fป*d
                          for tht ui< of both Jejcob and Ttitl* ซซi)jritl,  Th* •itf*MM initotropr of tht tqwlftr cm
                          bt Illuitntltf by tht l*rr/ป ring* Of lrซnmUlt*1tUt ullna thป two Mtfiwti, Oowbtt-Fon

                          My I Ho provldi ilatUr rttulti.*
216     CONTAMINATED AQUIFER CONTROL

-------
CONCLUSIONS

  The pilot scale treatment system can be evaluated for fabrication
time, mechanical performance and treatment efficiency. Approxi-
mately 4 hr with a two-man crew were required for field assembly
of the mobile treatment system. The minimization of field setup
time can help assure rapid mobilization and readiness in situations
where a time critical response is necessary. Mechanically, the system
performed as designed with one exception. During the last phase
of the St. Peter pumping test, a relay switch burned out in the final
discharge pump. Turbulence in Tank 4 created rapid on-off
switching of the float and burned out the relay. The turbulence
was corrected prior to the upper aquifer test. Modifications of the
switching system are under consideration.
  Treatment system efficiency can be evaluated by the results of
the Photovac screening until lab results are available. Treatment
efficiencies of greater than 99% for volatile organic contaminants
were achieved during the Galena-Platteville testing. The system
functioned mechanically, and from a treatability standpoint, as
designed.
  The results of the aquifer response testing provide valuable
insight into the feasibility of a groundwater pump-and-treat system
to remediate the aquifers of concern at the Byron Johnson Salvage
Yard. The technical decision to implement a purge system is bas-
ed primarily on hydraulic conductivities of the lithologies. A
preliminary analysis of the distribution of hydraulic conductivities
for the Galena-Platteville dolomites based on slug tests perform-
ed by Camp Dresser & McKee, USGS and U.S. EPA indicates that
the pumping well is indicative of the lower scale of hydraulic con-
ductivities found in the area.(3'6) As a result, the pumping rate us-
ed during aquifer response testing may be considered the minimum
which could be  designed for a purge system. For instance, max-
imum pumping rates would probably increase as wells are placed
closer to primary fracture systems. Maximum pumping rates will
obviously impact remediation costs and therefore the ultimate selec-
tion  of remedy  for the groundwater system.
  If a purge system is implemented  to intercept  all groundwater
flow emanating from the Salvage Yard, siting of the pumping wells
becomes the critical design  criterion. Well placement is dependent
upon maximum pumping rates and drawdowns. Because of the
anisotropic nature of the fractured upper aquifer system, maximum
pumping rates also  will be  dependent upon well placement.
Drawdowns probably will not be uniform or concentric but will
be ellipsoidal, and this also will impact well placement. Traditional
well siting rationales used in homogeneous aquifers, such as posi-
tioning wells along site boundaries, are not applicable for the
Salvage Yard. A purge system consisting of interceptor wells placed
perpendicular to both fracture systems and preferred orientations
of flow would be more effective.

ACKNOWLEDGEMENTS
  The authors would like to express appreciation to Barbara Ryan,
Robert Kay and David Olson  of the Illinois District of the USGS
for their review of,  and comments to, this manuscript. Additional
thanks are due to all of the personnel who provided their excellent
assistance in the field and office.

DISCLAIMER
  The information presented in this manuscript has not been sub-
jected to U.S. EPA review and,  therefore, does not necessarily
reflect the views of the Agency. No official endorsement should
be inferred.

REFERENCES

1.  Piskin.R.,  "Report on Toxic Wastes at Byron Salvage Yard, "for Illinois
   Environmental Protection Agency, 1976.
2.  Sargent & Lundy,  Inc., and Dames and Moore, Inc.,  "Fault Specific
   Geotechnical Investigations, Byron Station," Engineering Report, 1975.
3.  InteragencyAgreementsDW14932325-01-OandDW14932325-01-l, 1987.
4.  Kay, R.T., Ryan, B.J., Mears, E.J. and Yeskis, D.J., "Hydrogeology
   of the Byron/Johnson Salvage Yard Superfund Site, Byron, Illinois,"
   ASCE Water Res  Symp, 1987.
5.  Lohman, S.W., "Ground-Water Hydraulics," U.S. Geological Survey
   Professional Paper 708, 1979.
6.  Camp  Dresser &  McKee, "Technical Memorandum for Remedial
   Investigation Field  Activities for the Byron Johnson Salvage Yard Site,"
   for U.S. EPA, Work Assignment # 126-5L92.3, June 1987.
                                                                              CONTAMINATED AQUIFER CONTROL    217

-------
                Groundwater Treatment  for  Mixed  Contaminants:
                                        Processes  and  Facilities

                                             Thomas M.  Sanders,  P.E.
                                                   HDR Engineers
                                       Industrial/Hazardous Waste Division
                                                  Omaha, Nebraska
ABSTRACT
  The remediation of major hazardous-waste sites will require the
extraction and treatment of contaminated ground water. In recent
years, this treatment process has become more common; however,
the treatment frequently requirements  have been limited to air
stripping of volatile organics. As  more complicated sites are
approached, the required treatment operations increase in com-
plexity. Mixed-contaminant groundwater requiring treatment for
multiple classifications of pollutants will challenge the designers
of upcoming remedial actions. The case history described in this
paper outlines the design of a groundwater treatment plant con-
figured to treat volatile organics, heavy metals and phenols. Specific
topics discussed in this paper include: influent characteristics, dis-
charge options and limitations, treatability studies, process evalu-
ation and design; and facility development and characteristics.

INTRODUCTION
  Groundwater contamination may be out-of-sight, but it is not
out-of-mind to regulators and companies responsible for today's
contaminated groundwater cleanup projects because groundwater
is an increasingly important water resource. Groundwater con-
sumption in the U.S. has risen from 30 billion gallons/day in 1950
to 100 billion gallons per day in the early 1980s. Groundwater pro-
tection has become an increasingly important  issue as evidenced
by the following actions:

• The U.S. EPA created a  new Office of Groundwater  Protec-
  tion in April 1984
• In September 1984, the U.S. EPA formalized a groundwater pro-
  tection strategy
• The Safe Drinking Water Act was reauthorized in 1986 with new
  groundwater protection provisions
• The RCRA Amendments require control and monitoring pro-
  visions for underground storage tank systems
• The National Groundwater Policy Forum, a coalition of regula-
  tory,  environmental,  advocate  and  industry personnel, has
  recommended establishment of  a national  groundwater pro-
  tection goal.
• Expansion of laws and  regulations  governing groundwater
  cleanup and quality, protection is expected. The groundwater
  cleanup requirements outlined is this paper's case history are
  indicative of the growing expectations in this field.

  This paper's case history project resulted from a NPL site remedi-
ation and cleanup. The site operated as an industrial waste recycling
facility from the early 1960s through the early  1980s. The facility
accepted a wide variety of industrial waste materials:
  Electroplating solutions and sludges
  Pesticides and herbicides
  Waste oils and solvents
  Spent acid and caustic solutions
  Battery chips
  Flue dust
  Aluminum slag

  The facility operated over 10 to 15 acres and included labora-
tory, solvent recycling, drum storage, bulk liquid and dry material
storage and lagoon storage operations.
  Initial site  environmental investigations concluded that, in
addition to considerable surface feature cleanup, subsurface soil
and groundwater remediation would be required. The engineering
and science activities involved in assessing the site's contamina-
tion, designing remedial plans  and implementing the cleanup
encompass many disciplines. This paper focuses on the treatment
unit operations and the required treatment facility to accomplish
the contaminated groundwater treatment.

INFLUENT  CHARACTERISTICS
  The overall site remediation and cleanup includes the extraction
of groundwater from two distinct areas. Considerable subsurface
exploration and groundwater modeling was executed to develop
an  extraction  well  field plan. Key groundwater  extraction
parameters that had considerable impact on the treatment facility
development included:
• Contaminated groundwater existed from 5 to 40 ft. below grade
  in extremely heterogeneous geologic strata
• The area of groundwater extraction encompassed 10 to 15 acres
  and was recovered using more than 150 individual well points
• Considerable groundwater sampling/analysis had  been con-
  ducted; however, due to the heterogeneous geohydrology and
  ongoing cleanup activities such as specific, waste excavations,
  establishing a design basis contaminant profile was difficult

  The  project's  groundwater  treatment strategy required  the
independent  treatment of two streams. One stream, called the
primary stream in this paper, had a high flow rate and contained
highly mixed  contaminants. The secondary stream had a lower flow
rate and contained a very narrow profile of priority pollutants plus
significant levels  of inorganics such as iron and manganese.
   Influent characterization has a very  significant impact on
proposed treatment operations, on anticipated capital and opera-
tional costs and on confidence of meeting required discharge limits.
Figures 1 and 2 depict the design-basis influent profiles established
for the primary and secondary streams. Key project goals during
218    CONTAMINATED AQUIFER CONTROL

-------
the establishment of these profiles were:

• Reasonable interpretation  of highly scattered  groundwater
  quality data collected over several years from varying subsurface
  strata
• Allowance for the variations expected due to very heterogeneous
  geohydrology and the potential for selective usage of the large
  number of wellpoints
• Selection of average and maximum contaminant estimates that
  reasonably bracket the actual population of values, yet limit
  potential plant overdesign (and the associated cost) to the furthest
  extent possible

  With these goals, the primary stream was characterized as con-
taining three contaminant classifications: heavy metals/inorganics,
volatile organic compounds, and phenol and phenol compounds.
The design-basis contaminant mass loadings and concentration
levels are summarized in Figure 1. The secondary stream charac-
terization is  less complex, and the design-basis profile is shown
in Figure 2.  This stream's significant contaminants are volatile
priority  pollutant organics and  the inorganics are not priority
pollutants.
    -4-6 Ibs./lOOO gal. avg,
     20-30 Ibs/l000 g:
            3-4 IbsJIOOOgal. avg
            15-20 Ibs./lOOO gal max.
   Inorganics
   Iron
   Aluminum
   Manganese
   Zinc
   Nickle
   Chrome
   Cadmium
   Copper
   Others
- 1-2 Ibs./lOOO gal. avg
 15-20 lbs./1000 gal. max.
living*}
200 mg/l
 90 mg/l
100 mg/l
100 mg/l
 15 mg/l
  2 mg/l
 1.5 mg/l
 1 0 mg/l
10 0 mg/l
Volatile Organic
Compounds
Chloroform
1.1 Dichloroethana
Methylene Chloride
Trans 1 ,2 Dichloroethem
1,1,1 Trichloroethana
Trichloroethene
-Others—
\
(iv.ng.)
4 mg/l
4 mg/l
1 50 mg/l
40 mg/l
60 mg/l
40 mg/l
50 mg/l
Phenol and
Phenol Compounds
Phenol
Phenol Compounds
2-nitro, 4-nitro
chlorinated
2,4 -dimethyl
              72 mg/l
              2 mg/l
                            Figure 1
                Primary Stream Contaminant Profile
        - 0.2S-O.SO lbs./1000 gal. avg.
         2.0-4.0  lbs./1000gal. max.
        norganlc
        Compounds
        Manganese
        Iron
                       0.1-0.2 lbs./1000gal, avg.
                       1.0-2.0 lbs./1000 gal. max.
                      Volatile Organics
                                        (•virigi)
                      Trans 1,2 Dichloroethylene  10 mg/l
                      Others               <1 mg/l
                            Figure 2
               Secondary Stream Contaminant Profile

  Two additional influent contaminant parameters played a key
role in shaping the treatment facility design:

• The assumptions regarding how  maximum influent loadings
  would occur and how the treatment facility would react
• How the contaminant concentrations could be expected to trend
  over the long-term groundwater cleanup activity

  Figure 3 summarizes the project's assumptions regarding these
issues. First,  the  overall composite mass of contaminants  will
                                                       decrease over the 6 to 10 years of anticipated groundwater extrac-
                                                       tion and treatment. Second, the design-basis maximum (peak) con-
                                                       taminant concentrations are assumed to be due to the heterogeneous
                                                       nature of the geohydrology and the waste deposition. These peaks
                                                       could, therefore, be leveled by selective use of the individual well-
                                                       points. For this reason, the maximum contaminant loadings
                                                       indicated previously by Figure 1 and 2 are considered as short, term
                                                       spikes to the treatment plant.
                                                           1000
                                                                                                 Remarks'.
                                                                                                  Influent contaminants decline as
                                                                                                     groundwater cleanup occurs

                                                                                                  Contaminant concentration spikes
                                                                                                     limited to short-term events
                                                                                                     by welHiekJ control
                                                                                  1     23456789    10
                                                                                  Groundwater Pumping  Period, Years
                                                                                               Figure 3
                                                                                   Long Term Contamination Behavior
  The establishment of treatment flowrates was a technical issue
regarding soil/ground, water cleanup levels and desired project life-
time. Groundwater sweep, contaminant flushing, pore volume turn-
over and many other factors controlled flowrate selection. These
issues are outside this paper's scope. The primary stream treat-
ment flowrate was set at 200 gal/per minute; the secondary stream
treatment flowrate was set at 45 gal/per min.
                        DISCHARGE OPTIONS AND LIMITATIONS
                          The treatment facility was designed for discharge of the two
                        streams to alternative locations. Discharge options for the plant
                        included: (1) a publicly-owned treatment works (POTW), (2) direct
                        discharge to a stream and (3) infiltration back to the area of ground-
                        water extraction. Treatment-level requirements  for POTW dis-
                        charge were similar but somewhat more stringent than required
                        for pretreatment discharge of metal finishing and electroplating
                        categorical wastes. Treatment-level requirements for reinfiltration
                        on-site were nearly an order-of-magnitude more stringent than for
                        POTW  disposal. Direct-discharge  to the stream  requires con-
                        taminant removals dictated by consideration of stream background
                        quality and proposed safe drinking water parameters.
                                                        TREATABILITY STUDIES
                                                          The expected influent to this project's water treatment facility
                                                        contains contaminants commonly found in industrial treatment
                                                        trade wastes. The heavy metals, volatile organics and phenols are
                                                        common to metal finishing, electronics, automotive/aerospace and
                                                        the chemical-process industries. Although treatment approaches
                                                        to these industrial wastes sometimes can be developed based on
                                                        a theoretical  and/or previous experience basis, this project's
                                                        groundwater contamination was deemed too complex to rely on
                                                        a "desk-top" solution.
                                                          Two independent treatability studies were conducted approxi-
                                                        mately 6  months apart  during the  development of  the cleanup
                                                        strategy and treatment plant design. Key treatability study goals

                                                                   CONTAMINATED AQUIFER CONTROL    219

-------
and findings are tabulated below:
  Coals
                            Findings and Remarks
  Determine Ability to Remove Heavy   TreatablHty Indicated either How or sodium
  Metals ป1th Conventional Hydroilde  hydroxide addition mulled In good metals
  Precipitation                 removal; a tingle pH value workable even though
                            metals have varying theoretical minimum
                            solubility pH values
  Ability to Chemically Oxidize
  Phenol using Hydrogen Peroxide
                          Good treatment results obtained ซlth process
                          similar to most references; Intense black color
                          In treated water

                          Metals removal essentially Identical before and
                          after oxidation; black color removed efficiently
                          during metals precipitation and cIlrI Meat Ion
  Determine Potential Interference
  of Hydroxide Metal Precipitation
  by Hydrogen Peroxide Oxidation;
  and ability of Hydroxide
  Precipitation Process to remove
  black color

  Develop process parameters for air  Removal rates Identified and similar to most
  stripping volatile organlcs from   references
  •Ixed-contamlnant groundwater
Predict sludge production rates
and dewaterlng characteristics

Identify metal chelates present
In groundwater; determine ability
of processes to break chelates
                            Necessary parameters deflnvd
                            No significant clwlates present; clinical
                            oxidation operation capable of breaking
                            limited chelated metals prior to hydroxldi
                            precipitation
   In these two studies, which can only be summarized in of this
 paper, alternative unit operations were investigated. The results
 led to establishing design-basis treatment parameters with reason-
 able confidence.

 PROCESS EVALUATION AND DESIGN
   The treatability studies provided the basis on which to design
 specific process unit operations. In addition, other processes such
 as filtration and carbon adsorption were considered. Two process
 flow diagrams were developed and systems were designed to provide
 treatment for the primary and secondary streams. These two flow
 diagrams, shown as Figures 4 and 5, are the result of evaluations
 using the known influent characteristics, required effluent criteria
 and unit-operation behavior predicted by the treatability studies.
                             Figure 4
               Process Flow Diagram Primary Stream
  The primary stream treatment system uses a series of physical-
chemical unit operations to treat the contaminated groundwater
containing contaminants requiring several treatment operations:

• Hexavalent-chrome reduction
• Phenol chemical oxidation
• Volatile organics  air stripping
• Heavy metals precipitation

  These treatment  requirements are  met by the  process  flow
diagram shown in  Figure  4.  Specific unit  operations and  key
functions:

• Surge Tank. The  well field extraction system discharges into a
  surge tank. This equalization unit operation accomplishes:
  -  Flow equalization between the  well field  and the treatment
     process
  -  Limited pollutant-loading equalization
  -  Limited iron oxidation/removal with sludge pumping to the
     thickener
• Chrome Treatment. Dedicated unit operations  (and  process
  vessels) are provided for hexavalent chrome reduction. Key ele-
  ments include:
     Hexavalent-chrome reduced to the trivalent state under acidic
     conditions using sodium  bisulfite
     Chrome process vessel exhausted to  building exterior
     Phenol Oxidation. A dedicated process vessel is provided for
     phenol oxidation using hydrogen peroxide.
• Volatile Organic Air Stripping. A countercurrent packed tower
  unit operation is  provided for removal of volatile  organics:
  -  pH adjustment to enhances removal efficiencies  and/or for
     packing cleaning
  -  Air exhaust and stripped contaminants  are  discharged to
     emission abatement equipment
• Heavy Metals Precipitation/Clarification.  The inorganic metal
  contaminants  are treated   in a  traditional metal  hydroxide
  operation:
  -  Individual process vessels are provided  for pH adjustment,
     flocculation and clarification
  -  Clarification is provided  by an inclined plate unit
  -  Operational flexibility provided by  rqultiple  flocculent-aid
     chemical additions
• Sludge Dewatering. Wastewater contaminants removed as solids
  from the clarification process are  dewatered for transport and
  disposal off-site:
     The clarifier underflow (0.1 to 1.0% solids) is transferred to
     a thickener vessel
     Thickened (1.0  to 3.0% solids) sludge is pumped to a recessed
     plate  filter press  for liquid-solids separation;  30% to 45%
     solids content sludge cake is produced
                                    Hypochlorll*
                                                         Dltcharg*
                    Gr..ntind  Fllt.n       A|r str|pp.,r
                                          Tower
                              Figure 5
               Process Flow Diagram Secondary Stream
                                                                      The secondary stream treatment system will also use physical
                                                                    chemical unit operations to treat the contaminated groundwater.
                                                                    The influent contains contaminants requiring:
                                                                      •  Iron removal
                                                                      •  Volatile organics air stripping
                                                                       These  treatment requirements  are met  by the  process  flow
                                                                    diagram shown in Figure 5. Iron and manganese are present in the
                                                                    secondary stream groundwater. These two impurities do not require
                                                                    removal due to environmental concerns; however, their presence
                                                                    would impair  other treatment unit operations:

                                                                    •  Precipitants formed from the oxidation of these materials can
                                                                       form  scales and accumulations
                                                                    •  Iron and manganese bacteria can form objectionable slimes and
                                                                       furbecles in process piping and equipment
220    CONTAMINATED AQUIFER CONTROL

-------
  Alternative iron !/s manganese treatment methods fall into four
categories:

• Oxidation-Precipitation
• Direct-Ion Removal
• Sequestering
• Disinfection

  The secondary stream treatment system uses a type of direct-
ion removal known as zeolite filtration. These units use a bed of
natural green sand (zeolite) coated with manganese dioxide. As the
water flows through the bed, the dissolved iron and manganese
are directly oxidized and removed. The bed is regenerated by in-
termittent backwashing and flushing with a potassium perman-
ganate solution. These units include an anthracite filter layer over
the green sand to reduce clogging and increase filter run times.
  Key factors that lead to the selection of this process include:

  Moderate capital cost
  Moderate operational cost
  Reliable contaminant removal
  No need to repump water to  downstream operations
  Ability to fine-tune iron/manganese removal levels using bypass
  controls

  A countercurrent packed tower unit operation is provided for
removal of volatile organics. Key elements include:

• Chemical feed of  chlorine  solution  allows intermittent or
continuous disinfection
• Air exhaust and air stripped  contaminants are discharged to
emission abatement equipment

FACILITY DEVELOPMENT AND CHARACTERISTICS
  The groundwater treatment facility consists of treatment equip-
ment as well  as building  and site elements.
  Figure 6 and 7 show site arrangement concepts of alternative
treatment facilities that were developed in the early phases of
design. Figures 8 and 9 show building cutaway views of alternate
designs.
 \
                                                Site Arrangement
                            Figure 6

 SPECIAL STUDIES AND CONSIDERATIONS
   As this groundwater treatment facility was designed, a number
 of technical and regulatory issues surfaced and were acknowledged
 with appropriate solutions. One specific item required significant
 project response in order to maintain progress and is discussed in
 this section.
                                                 Site Arrangement
                           Figure 7
                                                                                                             Facility Configuration
                                                                                            Figure 8
                                                                                                                  Ficlllty Configuration
                                                                                             Figure 9
  The process development activities leading to the proposed flow
diagram and facility outlined in the previous sections were premised
on a key assumption: volatile organic compounds removed from
the contaminated  groundwater  could be  discharged  to  the
atmosphere. This assumption  later proved to  be  invalid  and
required plant modifications.
  Under average conditions, the originally planned air stripping
process would have discharged 500 Ibs/day of volatile organics to
the atmosphere. Peak discharges approaching 3000 Ibs/day were
predicted.
                                                                              CONTAMINATED AQUIFER CONTROL     221

-------
   A number of emission control alternatives were evaluated:          - hot gas stripping of the VOCs from the activated carbon had
   ,,  ,      .           ,, .          ..                            been determined to meet the project's goals.
   Carbon adsorption - off-site regeneration                                                   r       ฐ
   Carbon adsorption - steam stripped                              CONCLUSION
   Carbon adsorption - hot gas stripped                              _.   .   ,         ,         ,  ,            ,    ,     .   .
   Wet scrubbing   non aqueous solutions                            Thc development of a successful treatment plant for mixed-
   cซi,,oซ. ,,~~~~~~ A-   ป;„„                                      contaminant groundwater requires careful process development and
   fa
-------
                    Effective  Startup  and  Operational  Procedures
                          For Groundwater Remediation  Systems
                                                   David W.  Hale
                                                    Marc J. Dent
                                            David G.  Van Arnam, P.E.
                                          O'Brien & Gere Engineers, Inc.
                                                Syracuse, New York
ABSTRACT
  Several groundwater  remediation systems employing an  air
stripper for the removal of volatile organic compounds from a con-
taminated aquifer were selected for study. Startup,  sampling,
monitoring, system components and operating procedures from
four groundwater remediation systems were evaluated. State and
federal regulatory permitting requirements for system operations
were detailed. In addition, design considerations implemented to
prevent problems during startup and operation of the treatment
systems were addressed.

INTRODUCTION
  The discovery of contaminated compounds in municipal ground-
water supplies is becoming a frequent headline in many newspapers
across the United States. Implementation of effective groundwater
treatment systems is crucial for remediating these contaminated
aquifers. Where contamination of drinking water supplies is caused
by volatile organic compounds (VOCs), a treatment system em-
ploying an air stripper frequently is selected for the removal of
these contaminants from the groundwater.
  The effective operation of the air stripper, as well as other com-
ponents of the groundwater remediation system, is extremely im-
portant to prevent the further migration of contaminants within
the aquifer. Just as critical is the initial startup and monitoring
of the remediation system. Contaminant concentration level fluc-
tuations, electrical and mechanical equipment malfunctions require
that prudent startup procedures be implemented to prevent the dis-
charge of contaminants in excess of permitted levels or the release
of untreated groundwater to an area not previously contaminated.
  This study selected four groundwater remediation systems that
employed an air stripper for removal of VOCs in contaminated
groundwater aquifers. Regulatory requirements, system compo-
nents, startup and operating and monitoring procedures for  the
groundwater remediation systems are summarized in this study.
Various problems which can be avoided also are addressed.

BACKGROUND
  Four groundwater remediation systems were selected for discus-
sion in this paper. All the projects were initiated as a result of
federal and/or state regulatory agencies testing drinking water
supply wells as a result of user complaints and finding low levels
of VOCs present in the water supply. Typical VOCs discovered
by this testing included trichloroethene,  1,1,1-trichloroethane,
1,1-dichloroethane, t-l,2-dichloroethene and vinyl chloride. The
discovery of these contaminants prompted a groundwater remedi-
ation program to locate, quantify and remediate the contaminated
 plume in order to minimize the future impact on the aquifer.

 REGULATORY REQUIREMENTS AND PERMITS
   During all phases of the groundwater remediation projects, state
 and federal agencies were involved with the review and implemen-
 tation of the treatment systems. Their involvement was critical to
 insure that the design and operation of the remedial system did
 not violate permit requirements. For all four projects, regulatory
 agencies required both air emission and effluent discharge permits.
   The air emission permit greatly impacts the design of the remedi-
 ation system by establishing a maximum yearly discharge level for
 VOCs. Information regarding the anticipated concentrations of
 contaminants being emitted must be reviewed to determine if the
 treatment  process will violate maximum yearly discharge levels.
 If there is the  potential to exceed these maximum levels, alternate
 treatment  technologies or additional treatment of air emissions
 from the air stripper must be employed.
   During startup, violations of the maximum yearly discharge level
 is not a concern due to the permits being based on a Ib/year basis
 with no daily maximum. All groundwater systems had an initially
 high concentration of contaminant air emissions due to the recovery
 wells being installed in areas of high groundwater contamination.
 This initially high concentration is of concern to nearby populated
 areas and  on-site personnel. Precautions should be taken during
 startup to insure favorable atmospheric conditions to prevent an
 adverse impact to the nearby population. As a precaution, close
 monitoring of air emission should occur until the system reaches
 equilibrium.
   During startup, compliance with effluent discharge permits was
 critical to avoid civil and/or criminal penalties which could result
 from permit violations. Close monitoring of discharge levels for
 pH, flow and total VOCs insured these permit limitations were
 not violated.

REMEDIATION SYSTEM COMPONENTS
  The purpose of the individual system components is to func-
tion as a single unit to prevent future migration of contaminants
by collecting, transporting and treating the contaminated ground-
water. The major components common to  all the groundwater
remediation systems evaluated in this study included recovery wells,
pumps, vaults, collection vault, air stripper, blower, pipelines and
electrical controls (Figure 1).
  Data relating to the sizes and capacities of the individual com-
ponents of the four groundwater remediation systems selected for
discussion  are presented in Table 1.  The  following is a  general
description of the various components.

           CONTAMINATED AQUIFER CONTROL    223

-------
                  CONTAMINATED  GROUNDWATER
                                i
                         RECOVERY WELLS,
                         PUMPS, 8 VAULTS
                               I
                            COLLECTION
                               VAULT
  DISCHARGE TO
   ATMOSPHERE   _
   (AIR EMMISIONS
   PERMIT- AS
   REQUIRED )
AIR STRIPPER



BLOWER

                         DISCHARGE  TO
                         SURFACE WATER
                         (NPDES  OR SPDES
                         PERMIT REQUIRED)

                           Figure 1
              Typical System Components Schematic
Recovery Wells, Pumps and Vaults
  The groundwater recovery wells consist of a steel casing and a
slotted screen section installed at  a predetermined depth in the
aquifer to effectively collect the contaminated groundwater. All
recovery wells contain submersible pumps which discharge the con-
taminated groundwater, via pipelines, to the treatment system. In
order to impact a populated area as little as  possible, all electrical
controls and piping were installed in underground concrete vaults.
The submersible pumps in the recovery wells were selected to deliver
a flow rate greater than  the required design flow. By adjusting the
butterfly valve located in the vault, the flow from each recovery
well could be adjusted to fine tune the system. Flow sensors were
located in the vaults to help regulate the flow from each recovery
well. The check valve in the vault prevented cross-contamination
of the recovery wells in  the event of a mechanical failure or shut-
down of an individual  pump in the system.  A sample tap was
provided in the vault to  collect groundwater samples and evaluate
the contaminants at each well location. A groundwater level sensor
was installed in each recovery well to regulate the on/off cycling
of the pumps and provide an accurate depth to groundwater reading
(Figure 2).


CHECK





WELL
SLEEVE TYPE
COUPLING -y






\ N
\ r

>x



L3
CASING 	 >
GROUNOWATER
SENSOR CABLE-
r-FLOW SENSOR
FINISHED GRADE "V
	 TIIWB?


$s
rF^ป 	 JT —
1
1 r-SUMP
L *-


c

„
/SAMPLE TAP
COLLECTION VAULT
— BUTTERFLY VALVE


P- :-" ,' -1


1




i
jfj 	 RISER PIPE
(FROM WELL PUMP)
                                                                  LEVEL  ELECTRODE
                                                                                            -VERTICAL  SUBMERSIBLE  PUMP
                                                                       -SLOTTED SCREEN
                                                                                             Figure 2
                                                                               Typical Recovery Well, Pump and Vault
                                            Collection Vault
                                              The collection vault receives contaminated groundwater from
                                            the various recovery wells and acts as an equalization basin for
                                            the air stripper. This provides a steady flow of groundwater to the
                                            air stripper while minimizing the cycling of the submersible pumps.
                                            These submersible pumps have sufficient capacity to individually
                                            deliver the total system's flow in the event of one pump's failure.
                                              For these projects, the collection vaults were constructed of pre-
                                            cast concrete manholes with various diameters and depths. Each
                                            vault  contained a water level sensor to regulate the on and off cycles
                                            of the submersible pumps (Figure  3).

                                            Air Stripper
                                              The air stripper provides the mechanism for VOC removal from
                                            the contaminated groundwater. The pumps located within the col-
                                            lection vault discharge the contaminated groundwater to the top
                                            of the stripper. The contaminated groundwater travels downward
                                            over packing material, which has an extremely large surface area,
                                            towards the stripper outlet at the base of the tower. At the same
                                            time, air from the blowers is forced upward through the packing
                                            material and comes in contact with the contaminated groundwater.
                                            VOCs are stripped from the contaminated groundwater as a result
                                            of this action.
                                              A flow meter and recorder were provided on the influent line
                                            of the air stripper to monitor and record the flow from the collec-
                                            tion vault. Sample taps were installed on the influent and effluent
                                            line of the air stripper (Figure 4).
  Remediation System

         A

         B

         C

         D
 No. of
Recovery
  Well*

    2

    1

    a

    3
                                                            Table 1
                                               Remediation System Components Data
                                                                       Air Stripper
 Capacity of   Collection Vault        	
Recovery Well      Pump
Pumps (gpm)  Capacity  (gpm)  Capacity (gpm)  Diameter (ft)

  200*100         NA            100           •

    5-50          100            100          2.5

   50-200          1500          1500          1.5

  700-1000         NA           2500           9
 Blower
Capacity
 Icfs)
Sites of
Pipeline
 Inches)
  Length
    of
Pipeline (ft)

   1000

   5200

   6900

   2500
 224    CONTAMINATED AQUIFER CONTROL

-------
                            SLEEVE  TYPE
                            COUPLING1TYP.)
BUTTERFLY VALVE (TYP)

    CHECK VALVE (TYP)
FORCE MAIN FROM
RECOVERY WELLS-i
                                             PUMP DISCHARGES
                                T
                                    8'0  PRECAST
                                    COLLECTION  VAULT
                            Figure 3
                Typical Collection Vault and Pumps
            STRIPPER
            INLET 	
     FLOW  METER -
     FROM
     COLLECTION
     VAULT -
   SAMPLE
   TAP  (TYP. )-i
                                           -PACKING  REMOVAL
                                            DOOR
                                           INSPECTION PORT
                                           -PACKING  FILL
                                            DOOR
                                           -PACKED  COLUMN
                                            AIR STRIPPER
                           Figure 4
                 Typical Air Stripper and Blower
  The following design considerations were implemented to avoid
problems with the electrical controls due to the potential for the
flooding of the vaults.

• In areas with a high groundwater table, a waterproof sealer was
  applied  to the outside walls of the vaults and all wall penetra-
  tions  were caulked watertight
• Vaults were located above the flood plain whenever possible
• Sump pits and pumps were installed in all vaults to provide an
  area for water  to drain and to allow for water removal

Collection Vault
  Startup  of the collection vault pumps and the individual recovery
well pumps were coordinated to prevent an overflow condition at
the collection vault. A check of the various valves and eqpipment
at the collection vaults was performed prior to startup. Tfye check
valves and butterfly valves were operated to check for binding,
and the butterfly valves were completely opened.
  The water level sensors in the collection vaults  were calibrated
and tested to verify that the collection vault pumps would auto-
matically start and stop as the water level fluctuated. The blower
and air  stripper were operated during this phase of the startup.
  After operating the system for a short period of time, the total
system was shut off. The automatic drain valve was inspected to
verify that it opened when the collection vault pumps shut off.  This
valve was provided to allow the pipeline to drain, thus preventing
freezing during the winter. The  vault was drained after  all the
recovery wells were tested to remove any stones that may have
entered  the vault during  construction.
  The following design considerations were implemented  during
the installation of the collection vault and pumps:

• Teflon seals were installed between the two concrete sections of
  the collection vault due to chemical  incompatibility  of the
  standard O-ring joint and the contaminated groundwater
• The inner walls of the collection vault were coated with a water-
  proof paint to prevent leaching of the contaminated groundwater
  through the concrete and into a previously uncontaminated  area
• Strainers were installed on the inlet of the pumps  to prevent
  stones from passing through the pump and plugging the distri-
  bution spray nozzles in the air stripper.

Air Stripper
  As previously mentioned, startup of the air stripper was coor-
dinated with  start-up of individual recovery well pumps and the
collection  vault pumps. In order to avoid discharging groundwater
that could violate the discharge  permit requirements during the
initial startup, flow from the air stripper was discharged to a  tem-
porary retaining pool rather than surface water. One of the blowers
was always operated during startup of the pumps. During opera-
tion of the stripper, the flow meter and recorder were checked for
proper operation.
  Inspection of the distribution spray nozzles and the packing
material during testing of the equipment was also  conducted. The
distribution nozzles located at the top of the stripper tower evenly
distribute groundwater across the surface of the  packing. In the
event that individual nozzles became plugged or the distance
between the nozzles and packing was incorrect, the spray pattern
could be significantly altered, thus causing "short circuiting" of
the groundwater through the air stripper. This could cause the VOC
removal efficiency of the system to be  significantly  reduced,
resulting in possible violation of the discharge permit requirements.
  To assist in the inspection of the nozzles and packing, a perma-
nent access ladder and platform were provided. Clear vision panels
were installed  over  the inspection ports, packing removal and
packing fill doors to view interior operations.

Blowers
  During startup of the  individual recovery well pumps, one of
the blowers was always operating to meet the regulatory require-

           CONTAMINATED  AQUIFER CONTROL     225

-------
ments for effluent discharge. A predetermined volume of air was
required  to accomplish this.
  Inlet ports on the air duct between the blowers and the air strip-
per were utilized to check the air velocity and determine the actual
volume of air delivered to the air stripper. Air pressure gauges were
visually inspected to monitor the pressure drop through the air strip-
per. Low pressure readings indicated an air leak in the system. The
individual blower dampers were checked to verify that the air was
directed to the air stripper and not through the back-up blowers.

Blower
  At the base of each air  stripper, two blowers were provided to
deliver air to the air stripper.  It is this air flow flowing counter-
current to the water which facilitated the removal of VOCs as it
passed through the packing material. Each blower had sufficient
capacity to deliver the total required air supply and therefore would
act as a backup unit if one blower failed. A pressure gauge and
inlet ports were installed on the air duct between the blowers and
the air stripper. Automatic dampers were provided in the ducts
from each blower to insure that air from an individual blower was
directed to  the  air stripper (Figure 4).

PipeUnes
  Pipelines were installed for the conveyance of groundwater to
and from the various system components. Various recovery well
locations required the installation of cased pipelines under city
streets, streams and railroads. To aid in  the disassembly and
removal of equipment and valves, sleeve-type couplings were in-
stalled at critical locations in the piping system.

Electrical Controls
  Electrical controls play a major role in the  startup and opera-
tion of the  remediation system. Various controls were provided
for automatic backup of certain equipment and to shut-down the
system in the event of equipment failure. This control system is
essential  to prevent the release of contaminated groundwater to
surface water or the ground. High level alarms in the collection
vault and air stripper and loss of air flow indication were provided
to automatically shut off all the recovery well pumps. An  auto-
matic dialer was supplied to alert the facility  operator whenever
there was a problem at the facility.

STARTUP PROCEDURES
  Design considerations and startup procedures were implemented
to provide a remediation system  that complied with regulatory
agency discharge requirements. These design considerations and
startup procedures are summarized below.

Recovery Wells, Pumps and Vaults
  Prior to startup of the individual recovery wells, a check of the
various components was performed. The butterfly valve and check
valve were operated to check for binding, and the butterfly valve
was opened approximately three quarters of the way. The flow
meter display was visually inspected to verify that the unit was oper-
ational, and the groundwater level sensor's on/off setting was
checked to insure proper  starting and stopping of the pumps.
  Only one recovery well  pump was started at a time. While  the
pump was running, the butterfly valve was operated to verify that
the flow meter responded to the changing conditions. Finally,  the
desired design flow rate was obtained by throttling the butterfly
valve. During startup sequence, the collection vault pumps, blower
and air stripper  were operating. By slightly oversizing the pumps,
the system was  designed to be operated with  the butterfly valve
in the throttled position. In the future,  this oversizing would allow
more groundwater to be pumped from areas of high levels of con-
tamination and  less groundwater to be pumped from areas which
had low  levels of contamination  to maintain maximum system
efficiency.

Pipeline
  Prior to  startup of  the remediation system, all pipes were

226    CONTAMINATED AQUIFER CONTROL
hydrostatically pressure tested to insure the integrity of the pipe
and the joints. For any pipe sections that failed, corrective measures
were taken to repair or replace the pipeline.
Electrical Controls
  As part of the electrical startup procedure, automatic operations
were checked to verify proper system operations and to verify that
the electrical control system would prevent an overflow condition
or discharge of untreated groundwater.
  The groundwater level sensor installed in each recovery weU was
tested to verify that the recovery well pumps would automatically
shut off in the event low groundwater conditions occurred. This
would prevent the pump from running dry and burning out.
  Testing  of the high level sensor located in  the collection vault
was performed to verify that all the recovery well pumps would
automatically shut off if the water level in the collection vault
reached a predetermined level prior to overflowing the vault. By
shutting off the recovery wells,  a spill of untreated groundwater
would be averted.
  A high level sensor in the air stripper sump was checked to verify
that all  the recovery  well pumps and the collection vault pumps
would automatically  shut off if the water level reached a specific
level. This control system prevented water from entering the air
ducts and  damaging the blowers.
  The proof of an air switch, located in the air duct between the
blowers and the air stripper, was tested to verify that the recovery
well pumps and the collection vault pumps would automatically
shut off in the event  of blower failure. By shutting off the entire
system, the discharge of untreated groundwater could be prevented.
  The automatic switch over from one collection vault pump to
the other or one blower to the other was checked to determine that
the backup unit would function. This provides system reliability
and prevents the system from automatically  shutting down.
  In the event the entire remediation  system shuts off due to a
power or equipment failure, an automatic dialer would call the
facility  operator. This notification allows the operator to take
appropriate measures to correct the situation in a timely manner.
The automatic dialer was tested under the various emergency con-
ditions to  verify that it responded accordingly.

SAMPLING AND MONITORING
  During the startup period, the treated groundwater that was col-
lected in the retaining pool was sampled an*analyzed. If the con-
taminant  levels  in   the treated groundwater  were  below  the
regulatory discharge levels, the treated groundwater was discharged
to the receiving surface water. If the contaminant levels were above
the discharge levels, it was recycled to the collection vault.
  Following startup and during the first 2 weeks of operation,
sampling and monitoring of the groundwater effluent was extremely
important to prevent violating the discharge permit requirements.
Samples were  collected from the sample taps provided on the
influent and effluent pipes of the air stripper and analyzed on a
daily basis during this period to develop a good data base for evalu-
ation. VOC removal efficiency was calculated from the analytical
results to  check  the  performance  of the air stripper.
  Samples were also collected from the  individual recovery well
vaults and analyzed to determine the contaminant concentration
levels.  This information was used to regulate the flow from the
individual recovery wells. After the initial 2-week period, samples
were collected on a  weekly and eventually a monthly basis, as
required by the discharge permit.

OPERATIONAL PROCEDURES
  A well-operated and maintained remediation system is very
important. Proper operating procedures and a regular maintenance
program will minimize costs and provide a system that will operate
in accordance with regulatory requirements. A  responsible remedi-
ation system operator, who is knowledgeable in equipment opera-
tions, maintenance and regulatory requirements, will enhance the

-------
potential for system optimization.
  Routine daily inspection programs were established to determine
if the groundwater remediation systems were operating at peak ef-
ficiency or were in need of maintenance. The major component
of these programs was the inspection report forms which were com-
pleted by the operator. These forms contained a check list of the
following inspection activities for the systems.

  Check all indicator  run lights (daily)
  Record individual pump operating hours (daily)
  Record individual pump flow readings (daily)
  Check recovery well vaults  for accumulation of water (daily)
  Operate all butterfly valves (weekly)
  Inspect blower  belts (weekly)
• Inspect air stripper nozzles and packing (weekly)
• Lubricate equipment (as recommended)
• Monitor groundwater analytical results
CONCLUSION
  Effective startup and operational procedures are essential for
a successful groundwater remediation system. Regulatory require-
ments play a major role in all aspects of design, startup and oper-
ation of the system. By considering the design aspects presented
in this study, problems that typically are encountered with this type
of system can be minimized. Finally, obtaining the maximum
efficiency from the system and minimizing costs requires a respon-
sible operator to implement a daily inspection and maintenance
program.
                                                                              CONTAMINATED AQUIFER CONTROL    227

-------
                    Development  of Methodologies  for  Evaluation
                                         of Well-Point  Systems
                                           T.Y. Richard  Lo, Ph.D., P.E.
                                                   Victor H.  Owens
                                                Ebasco Services, Inc.
                                                     Dallas, Texas
ABSTRACT
  A conceptual framework for quantifying the performance of
ground water remediation systems is presented. Example scenarios
are used to illustrate the quantification of waste isolation, fraction
of contaminant recovery and optimal well-point system configu-
ration. Current and future work are described discussing how the
final conceptual framework will be used as a basis to implement
rapid well-point evaluation  methodologies.

INTRODUCTION
  Well-point systems often have been used to achieve groundwater
remediation through interim waste source isolation and/or perma-
nent groundwater cleanup. Although groundwater remediation is
needed at most of the hazardous waste sites throughout the United
States, quantitative methods for evaluating well-point system per-
formance are poorly defined and usually are not implemented in
the planning phase. However, the high operating and maintenance
costs associated with long-term groundwater remediation indicate
a need for rapid quantitative evaluation of the well-point system
both before and after installation.
  A conceptual framework has been established to be used in the
development of a computer-aided design tool that formulates the
basis for rapid design and evaluation of complex well-point systems.
Long-term development of methodologies based on concepts for
quantifying remediation goals has been divided into three stages.
  In stage I, the means  to quantify the remedial objectives are
developed. For waste isolation, a concept of groundwater-waste
contact according to contact volume ratios combined with ground-
water migration control is adapted; and for groundwater cleanup,
a concept of fraction mass removal over time is adapted. For each
concept  is a means to quantify specific aspects of the cleanup
objective.
  Stage II involves development of methodologies to quantify the
performance of the well-point system in reaching the remedial
objectives. Fundamental geohydroiogic principals currently are
being combined with engineering design concepts in this stage of
development. Finally in Stage III, the methodology will be auto-
mated, creating an expert system to enable efficient use of this tech-
nology as a tool to optimize remedial design and evaluation.
  This paper presents the concepts developed and used in Stage  I.
Stage II and III development are still preliminary although initial
Phase II work currently  is being tested.

OBJECTIVE OF CONCEPTS
  One primary objective of this paper is to define a reasonable
set of basic concepts regarding well-point system performance as

228    CONTAMINATED AQUIFER  CONTROL
related to waste isolation and groundwater cleanup. These con-
cepts may appear fundamental to the field of professionals involved
in this work; however, this paper presents an organized set of
definitions for well-point system performance that can be uniformly
applied to all groundwater remediation efforts. Ultimately, our goal
is to develop methodologies, based on these concepts, in a manner
that allows  rapid and consistent evaluation of the optimal well-
point system for the remediation of groundwater contamination.
  In order to implement rapid evaluation for well-point systems,
the basic design goals of a well-point system have been generalized.
Our focus is on defining the performance of the well-point system
in terms that can be used in a consistent manner to converge to
optimal system performance. Initially, this performance is directly
proportional to the degree of waste isolation and/or contamina-
tion removal, irrespective of cost  consideration. However, the final
design of most well-point systems requires that optimal technical
performance criteria be relaxed to incorporate reasonable cost con-
straints. Therefore, the cost performance, while not directly
addressed here, is an integral part of the final  project objective.

DEFINITION OF REMEDIATION GOALS
  The need for clear, unambiguous goals for groundwater remedi-
ation is most obvious when redefined in terms of remedial design
criteria. When cleanup criteria are set  at background levels, the
well-point remediation goal is  clearly defined. Unfortunately,
remediation criteria are sometimes set without regard to true tech-
nical achievability and without regard to reasonable  cost con-
siderations.
  Therefore, in order to quantitatively define the scale upon which
remedial  action goals  can  be based,  two general concepts for
remediation have been considered: (1) degree or fraction of iso-
lation of the waste source such that leachate generation and migra-
tion can be  quantified and (2) fraction of waste mass removal per
unit time from the contaminated groundwater or unsaturated zone.
These concepts are considered in the following sections.

QUANTIFICATION OF WASTE ISOLATION
  As shown in Figure 1, waste material  can be submerged beneath
the standing water level such that leachate is produced from rain
percolation as well as  direct groundwater contact. In  order to
remove the waste source from direct groundwater contact, a simple
well-point system can be established as shown in  Figure 2. The
degree to which the well-point system is successful can be measured
based on the volume  of waste contacted before groundwater
extraction compared to volume  contacted after groundwater
extraction. Clearly, if the well-point system can  induce a localized

-------
water table surface completely below the waste source, then the
well-point system is 100% successful.  If rain infiltration can be
controlled by capping, again measured on a scale of zero to 100%
success when compared to no remediation, then the waste source
can  be completely  isolated such that leachate  generation is
completely halted.
                                                                 Infiltration after remediation
                                                                 Infiltration before remediation
o  _ j
 i
                                                                 Expansion rate of plump after remediation
                                                                 Expansion rate of plume before remediation
                                                                      Equation  1 provides a very simple means of quantifying the
                                                                    degree to which waste source isolation is successful on a scale of
                                                                    0.0 to 1.0.  The Isolation Factor provides a direct means of
                                                                    optimizing well-point systems intended to reduce the potential for
                                                                    leachate generation and/or migration

                                                                                                      \ Ground Wot*
                                                                                                     •~-\ Contominotion Zoni
                            Figure 1
       Generalized Waste Source in Contact with Groundwater
       (Rtiniicud Wattr)
                                                 _JH>ini>ctni Wqtซ)
                                                                                               Figure 3
                                                                     Groundwater Contamination Expansion Without Well-Point System
                                                                                                                     .Ground Wotlr
                                                                                                                            ion Zonf
                           Figure 2
       Generalized Waste Concept for Waste Source Isolation
                                                                                              •Clton Ground
                                                                                              Wottr Extraction Will
  However,  100% success in lowering the groundwater table
beneath the leachate source or 100% success of a cap may not be
technically or economically feasible. Considering these limitations,
it may be practical to slow the migration of contaminated ground-
water through a well-point system placed either up-gradient or as
an interceptor. Figures 3 and 4 illustrate the concept of limited con-
trol of contaminated groundwater migration. If the groundwater
containment is 100% successful, then no contaminated ground-
water will leave the influence of the extraction wells. However,
as with groundwater elevation changes and capping,  the economic
and technical feasibility of groundwater contamination control may
be less than 100% achievable. Therefore, the following definition
has been adopted to quantify the degree of waste isolation achieved
using an Isolation Factor (If):
       i  =
+ Rf)/3
                                                        (i)
where  R,, = 1 - —
       V, =
           Volume of waste material in contact with ground-
        , = water after steady state conditions with well-point
           system extraction

           Initial volume of waste material in contact with
        i ~ groundwater before well-point system extraction
                                                                          Figure 4
                                                   Groundwater Contamination Expansion With Well-Point System
                                                QUANTIFICATION OF WASTE REMOVAL
                                                  The second definition required to quantify remediation goals is
                                                based on removal of contamination from the groundwater. Typi-
                                                cal groundwater remediation by extraction is achieved by pump-
                                                and-treat methods, and performance is measured based on con-
                                                centrations of extracted water and monitor well samples. However,
                                                the remediation of groundwater can be a lengthy process spanning
                                                up to several years. In order to quantify the remedial objective such
                                                that well-point system optimization can be achieved, a concept of
                                                fraction mass removal over time has been adopted in terms of a
                                                Removal Factor (Rf):
     =5-
                                                                                                                        (2)
                                                where Mฃ = Total mass available for removal at T = 0
                                                     Re = Extraction rate of contaminant (time dependent)
                                                     T = Elapsed time of extraction through well-point system

                                                  Using this concept, the duration of containment recovery can
                                                be quantified  at any set recovery  fraction or vice versa.  The
                                                advantage of apply this concept is that planning phase activities

                                                           CONTAMINATED AQUIFER CONTROL     229

-------
can be based on more realistic project durations and remedial design
costs in combination with health risk evaluations. For example,
natural attenuation can play a very significant role in achieving
reduced health risks for many remedial designs. Using the con-
cept of fraction mass removal over time, the combined effects of
contaminant  extraction, decreased plume  migration rates and
extended biodegradation times can be considered simultaneously
with cost factors to produce the optimum well-point system at the
lowest possible present-value cost.


QUANTIFICATION OF OVERALL REMEDIATION GOALS
  Using the concepts of Isolation Factor and Removal Factor, it
is possible to define an independent variable scaling the remedia-
tion goal from 0.0 to  1.0, with  1.0 being 100% achievement of
remediation goal for a given duration of extraction through  the
well-point system(s). The defining equation for overall remedia-
tion efficiency  (E^ is simply the product of the component
factors:

Eo  =  !f Rf                                              (3)

E0 is the overall technical, environmental and cost efficiency of
the well-point system for a given project duration, the factors If
                                                                  and Rf are derived  ultimately from either  simple  or complex
                                                                  modeling efforts as required by the complexity of the remediation
                                                                  effort. Clearly, if If and Rf are individually optimized with respect
                                                                  to technical, environmental and cost considerations, then E0 will
                                                                  be optimized.  The degree to which modeling and/or data collec-
                                                                  tion are used  in determining If and R( will depend  on the indi-
                                                                  vidual demands  of the remedial design criteria.

                                                                  FUTURE WORK
                                                                    As stated previously, the concepts  presented in this paper are
                                                                  calculated  to  form  a framework upon which  quantitative for
                                                                  groundwater remediation will be based. The authors currently are
                                                                  developing analytical methods  for use  in determining the mass
                                                                  recovery of contaminant  from fully efficient extraction wells. A
                                                                  fully efficient extraction well is defined in these methodologies as
                                                                  a well that  recovers no uncontaminated groundwater while being
                                                                  used  to extract  contaminated  groundwater.  In  waste isolation
                                                                  studies, standard numerical models have been combined with the
                                                                  concept of  the Isolation Factor in order to produce optimum iso-
                                                                  lation strategies. Although not currently under development, the
                                                                  ultimate goal of these efforts is to prepare and implement an expert
                                                                  system useful in selecting the groundwater remediation strategy that
                                                                  minimizes environmental exposure, health risks, costs and duration
                                                                  of remediation.
230    CONTAMINATED AQUIFER CONTROL

-------
                 Comparison  of  Pollutant Fluxes  in  Saturated  and
               Unsaturated  Flows  Beneath Hazardous Waste  Sites

                                                  Lance R. Cooper
                                                     CH2M HILL
                                                  Atlanta, Georgia
                                              Frank L. Parker,  Ph.D.
                                                Vanderbilt University
                                                Nashville, Tennessee
ABSTRACT
  A study was conducted to compare predicted contaminant migra-
tion in groundwater regimes beneath hazardous wastes sites for
saturated and partially saturated flow conditions. Contaminant
migration beneath two waste disposal sites in Tennessee was pre-
dicted by modeling  subsurface systems using SUTRA (Saturated
Unsaturated TRAnsport), a United States Geological Survey model
written by Clifford I. Voss.
  Relative concentration  profiles  for saturated  and partially
saturated conditions at each site were plotted at various times in
the simulations. A factor of safety, or degree of conservatism, was
defined as the ratio  of the saturated steady-state concentration at
a distance from the site to the unsaturated steady, state concen-
tration at that distance. For both sites, this factor of safety was
less than two.

INTRODUCTION
  The practice of land disposal of hazardous materials has led to
contamination of groundwater resources in the vicinity of many
hazardous waste disposal sites. By predicting transport of chemicals
in groundwater, potential impacts of waste disposal sites on near-
by groundwater quality can be assessed. However, predicting the
movement of these contaminants can  be a very difficult task.
Simplifying assumptions are generally made to model groundwater
flow and chemical transport. One typical assumption  is that the
zone beneath the disposal facility is completely saturated.  This
assumption leads to a conservative solution to the problem  (i.e.,
a solution that yields faster contaminant migration), since most
waste sites have an area of unsaturated flow (the vadose zone)
beneath them.
  This study assesses the merit of including the unsaturated zone
in computer simulation of flow and transport beneath disposal
facilities. The sites  studied were the North Hollywood Dump in
Memphis, Tennessee and the Hardeman County, Tennessee dis-
posal site.

DESCRIPTION OF MODEL USED
  SUTRA (for Saturated Unsaturated TRAnsport) is a ground-
water transport model that allows simulation of either totally
saturated or combined saturated-unsaturated systems'. The model
employs a "hybrid finite-element and integrated finite difference"
solution  scheme. SUTRA allows simulation of variable  density
fluids (i.e., density changes with concentration), solute production
or decay and solute sorption onto soil matrix. The model may be
used for either area! or cross-sectional saturated simulation or cross-
sectional unsaturated simulation. Voss1  presents  a detailed
description of the model in the user's manual.

DATA REQUIREMENTS
  The data necessary to realistically represent a groundwater system
with a mathematical model include: values of physical parameters
distributed throughout the system; boundary and initial conditions;
fluid and solute sources and  sinks and unsaturated parameter
functions1'2.

Physical Parameters
  The physical input parameters needed to predict the fate of toxics
in groundwater include properties of porous media, fluid and
chemical contaminants1. One of the most complex issues in sub-
surface modeling is parameter evaluation. It is difficult to estimate
values of these parameters and even more difficult to surmise how
these properties will change in 5, 50, 500 or 5,000 years.
  Porous media parameter values include the intrinsic permeability
and porosity distributions throughout the system and the areal and
cross-sectional extent of geologic formations. These parameter dis-
tributions are difficult to define because data are gathered only
at a finite number of points. Important geologic formations, such
as sand lenses in a clay layer and fractures in a rock  formation,
may be missed.
  In addition to these parameters,  SUTRA requires that the longi-
tudinal and transverse (or maximum and minimum) dispersivities
be  included  in  order  to  simulate dispersion in the system.
Definitions of these parameters along with the functional relation-
ships are  discussed by Voss1.
  Fluid properties in groundwater systems are somewhat easier to
define. These properties include fluid viscosity, fluid compressi-
bility and fluid density, data which are given in most texts on fluid
mechanics or groundwater hydrology. The viscosity and density
of the fluid vary with both fluid temperature and solute concen-
tration. SUTRA  is designed  to  account for  the  variations of
viscosity with fluid temperature and the variation of density with
fluid temperature or solute concentration.
  Important chemical properties include water solubility, molecular
diffusivity and rate of natural decay, degradation or sorption.
Literature reports of solubility of a pollutant in water may vary
by  several orders  of magnitude. This point  is shown well by
Broshears3, who  found reported estimates of chlordane water
solubility ranging from 0.009 mg/1 to 1.85 mg/1. Reported mole-
cular diffusivities and rates of degradation also may show a high
degree of variability. Accepted practice is to use the most conser-
vative values for these parameters (i.e., the values that would yield
the fastest solute transport) in an effort to provide conservative
estimates.

Boundary and Initial Conditions
  Boundary and initial conditions must be specified in the  aquifer
to  model the groundwater  system.  These  conditions, when
combined with the governing equations, provide a unique solution
for the system4.  Three types of boundary conditions may be
specified:

• Specification of a constant or time-varying value for the depen-
  dent variable along the boundary (known as Dirichlet conditions)

           CONTAMINATED AQUIFER CONTROL     231

-------
• Specification of a constant or time-varying value for the flux
  across the boundary (known as Neumann conditions)
• A combination of Dirichlet and Neumann boundary conditions
  (known as Cauchy conditions)5.

  Boundary conditions may represent actual system boundaries,
as in the case of streams or hydraulic divides. In cases where no
well-defined natural boundary conditions  exist,  the  problem
domain normally is chosen large enough that the specified boundary
conditions will have little or no effect on the region of interest.
Fluid  and Solute Sources and Sinks
  Contamination of groundwater supplies from hazardous waste
landfills is the result of contaminated water leaching from the waste
site and entering the groundwater aquifer. Two types of leachate
may be generated if the waste is located above the water table. The
first type results from infiltrating precipitation that becomes con-
taminated as  it passes through the waste. The second  type of
leachate is any liquid waste that may have been disposed of in the
landfill.

Unsaturated Functions
  To  model the unsaturated zone it is necessary to provide the
functional relationship of parameters in this zone, including:

• Variation of moisture content (or saturation) with pressure head
• Variation of relative permeability with either moisture content
  or pressure head.
• Van Genuchten6 provides a discussion of a possible technique
  to estimate these relationships. The relationships are dependent
  upon soil parameters such as the saturated permeability, residual
  saturation (the point where the relative permeability is zero) and
  constants which depend on soil type. The programming neces-
  sary to represent these relationships is included in SUTRA1.

APPLICATION TO NORTH HOLLYWOOD DUMP

Site Description
  The North Hollywood Dump is located in the old meanders of
the Wolf River in Memphis, Tennessee. Uncontrolled dumping
occurred at the 28-hectare site from the mid-1930s to the mid-1950s.
From the mid-1950s to 1967, the dump was operated by the City
of Memphis for municipal and industrial waste disposal. The dump
is bordered by the Wolf River to the north, surface ponds to the
east and west and a residential neighborhood to the south3-7.
  The dump site overlies an alluvial deposit consisting of soils that
vary from clays to gravels. This alluvial layer consists primarily
of fine sands, silts and clays in irregular formations. Beneath this
alluvium is a fairly continuous  formation consisting of terrace
deposits that vary from 0 to 30 m  thick. These terrace deposits
are primarily sands and gravels. Beneath the terrace deposits is the
Jackson Formation, a 60 m thick, nearly impervious clay layer that
separates the upper sands and gravels from the Memphis Sand,
an aquifer that provides the primary drinking water source for the
city of Memphis.  Due to the presence of the confining Jackson
formation, the predominant transport of contaminants via ground-
water occurs in the alluvium and terrace deposits.
  The contaminants  of concern at the Hollywood Dump are
organochloropesticides and  pesticide-related  chemicals. Since
SUTRA models the transport of only one chemical species at a
time,  chlordane was chosen to represent  this group of pesticides.

Modeling  Saturated Flow and Transport
  The primary mechanism for  groundwater pollution from the
Hollywood Dump is  rainfall infiltration into the waste site. To
simulate transport from the site, the flow domain was discretized
(i.e.,  grid setup),  the physical parameters of the system were
estimated, the  fluid and solute source terms were estimated and
boundary and initial  conditions  were assigned.   The input
parameters needed to simulate contaminant transport at the North
Hollywood Dump (as well as the Hardeman County Disposal Site)
are presented in Table 1.

 232    CONTAMINATED AQUIFER CONTROL
                           Table 1
          Fluid and Soil Properties for SUTRA Simulation
       Parameter
  Permeability

  Dl*pซrslvity
    Baxinua
    BlnlBun

  Poroalty

  Solid* CoBpr***lblllty

  Solid* Oenilty

  Fluid Coeipreปซtblllty

  Fluid Demity

  Fluid Vlccoclty

  Chemical Specie* Modeled

  Water Solubllty

  Molecular Dlffiulvlty

  Decay Rate
                               north            Hardeซan
                            Hollywood Puปp       County 8itซ
  1x10    B*


     40 B
     4 B

     0.10

     0.0

  2650 kg/B3

4.4.10-10 .'
                        100 •
                         10 •

                        0.08

                         0.0

                     2650 kg/.3

                   4.4X10'10 „*/•
  1000 kg/B           1000 kg/B

 lxlO~3 kg/B-a       IxlO3 kg/B-*

   Chlordane     Curbon Tetrechlorl*

   1.85 Bg/1           800 Bg/1
  lxlO~9 B3/*

-2.29xlO~9 a"1
                     Ixio"9 B3/*
                        0.0
  The primary objective of this study was to compare saturated
and partially saturated fluxes and not to provide the "best" pre-
dictive simulation  of transport from the sites. Therefore, the
layered, heterogeneous and anisotropic system beneath the site was
approximated by a homogeneous medium.
  The finite element grid for the saturated flow condition repre-
sented a cross, section beneath the site that  was 400 m long and
8.5 m deep. The grid consisted of 20 horizontal elements (20 m
each) by 24 vertical elements (0.35 m each), for a total of 525 nodes
and 480 elements.
  Infiltration percolating to the water table was the major con-
taminant source for  the  groundwater system. This inflow was
assumed to occur at a constant  rate based upon average annual
infiltration. The rate of infiltration at the Hollywood Dump has
been estimated to be 0.29 m/yr3. The fluid inflow at those nodes
representing the dump was assigned a solute-concentration equal
to the water solubility of chlordane (1.85 mg/1). This method of
defining the solute source assumes that all water entering the landfill
becomes totally saturated with waste before entering the ground-
water. These sources were assumed to remain  constant through-
out the period of simulation.
  Boundary conditions were imposed upon the system as follows:

• A  constant  fluid flux  condition along the top boundary  to
  represent infiltration of rainfall into  the site
• A  no-flow condition along the lower boundary to simulate the
  relatively impermeable Jackson Formation
• Constant pressure conditions at the vertical system boundaries

  The values of the pressures along the vertical boundaries were
assigned as hydrostatic pressures (i.e., increasing linearly down-
ward due to the density  of the fluid).  To specify heads in this
manner, the value of the head at the two upper corner nodes must
be known.  The pressure at one node was assigned to be 210 m and
the value of the head at the other upper corner node was then calcu-
lated by using the prevailing gradient of 0.005 m/m.
  The simulation was conducted for a 10 year period, with time
steps of 1  day.
Modeling Unsaturaled/Saturated Flow  and Transport
  To simulate the unsaturated/saturated (hereafter referred to as
unsaturated) conditions at the Hollywood Dump, a few modifi-
cations  were needed for the input data file. The physical domain
of the problem was expanded to include the 3.5 m unsaturated zone

-------
beneath the site. It also was necessary to specify different boundary
conditions for the system as well as the parameters necessary for
the unsaturated functional relationships.
  The finite element grid for the unsaturated condition represented
a cross, section beneath the site that was 400 m long and 12 m deep.
The grid consisted of 20 horizontal elements (20 m each) and 24
vertical  elements (0.5 m each), for a total of 525 nodes and 480
elements. The saturated depth of the aquifer was 9.5m along one
vertical boundary and 7.5 m at the other. This represented a water
table slope of 0.005 m/m, the prevailing hydraulic gradient in the
aquifer.
  The physical input parameters in this unsaturated simulation are
identical to those used in the saturated simulation and were shown
in Table 1.
  Boundary and initial conditions  for the unsaturated simulation
were the same as those specified for saturated only flow with the
exception of the vertical constant head boundaries. The pressures
at the nodes  representing the water table along the vertical
boundaries were specified to be zero. The nodes beneath these were
assigned pressures equal to the hydrostatic pressure as was done
in the saturated simulation. The portions of the vertical boundaries
above the water table were specified  as no-flow boundaries.
   This simulation was conducted  for a 7-year period, with time
steps  of 1 day.

APPLICATION TO HARDEMAN COUNTY  SITE
Site Description
   The Hardeman County disposal site covers over 97 hectares in
Hardeman County, Tennessee. The actual disposal area is approxi-
mately  11  hectares8. The site  is located in a rural area and is
bordered on the east by Pugh Creek and on the north by Clover
Creek8. The site was operated by a chemical company from 1964
to  1972  as  an  industrial waste  dump  for pesticides  and
solvents3'8-9. The  leachate from the site has contaminated some
domestic wells in the area, leading to  remedial action at the site8.
This remedial action included placement of a clay  cap to reduce
infiltration and leachate.
   The disposal site is located in fluvial deposits that consist of
sands, silts and clays. Beneath this formation is the Claiborne
Formation, a mixture of sands and other materials. The Claiborne
Formation is underlain by the Wilcox Formation, and the two are
separated by an essentially impervious clay layer3.
   Heptachlor and carbon tetrachloride have  been detected in
monitoring wells downgradient from the disposal site. The chemi-
cal used in this simulation was carbon tetrachloride.

Modeling Saturated Flow and Transport
   As with the North Hollywood Dump, the primary mechanism
for contaminant transport from the Hardeman County disposal
site is that of infiltration passing through the waste site.  The
predominant groundwater flow direction is north, parallel to Pugh
Creek.
   The finite element grid for the saturated flow condition at the
Hardeman County site represented a cross-section beneath the site
that was 1000 m long and 4 m deep. The grid consisted of 10
horizontal  elements (100 m  each) by 8 vertical elements (0.5 m
each), for a total of 99 nodes and 80 elements.
   The input parameters that were needed to simulate contaminant
transport at the Hardeman County Disposal Site are presented in
Table 1 (along with input parameters  from the North Hollywood
Dump Site).
   Infiltration percolating to the water table was the major fluid
source for the groundwater system. As with the Hollywood Dump,
this inflow was assumed to occur at a constant rate based upon
average annual infiltration. The infiltration rate at the Hardeman
County site has been estimated to be 0.18 m/yr3. The fluid inflow
at those nodes representing the disposal site was assigned a solute
concentration equal to the water solubility of carbon tetrachloride
(800 mg/1). These fluid and solute sources were assumed to remain
constant throughout the period of simulation.
  Boundary conditions were imposed upon the system as follows:

• A constant fluid flux along the top boundary to represent infil-
  tration of rainfall into the site
• A no-flow condition along the lower boundary to simulate the
  relatively impermeable layer beneath the site
• Constant pressure conditions at the vertical system boundaries

  The  values of  pressures  along the vertical boundaries were
assigned as hydrostatic pressures after the value of head was defined
at the two upper corner nodes. The pressure at one upper node
was arbitrarily assigned to be 200 m,  and the value of the head
at the other upper  node was then calculated by using the prevailing
hydraulic gradient of 0.0034 m/m.
  The  simulation was conducted for  a 5-year period, with time
steps of 1 day.

Modeling Unsaturated Flow and Transport
  To simulate the  unsaturated conditions at the Hollywood Dump,
a few  modifications were needed for the  input data file.  The
physical domain of the problem was expanded to include the 15
m unsaturated zone beneath the site. It also was necessary to specify
different boundary  conditions for the  system  as well  as  the
parameters necessary for the unsaturated functional relationships.
  The  finite element grid for the unsaturated condition represented
a cross-section beneath the  site that was 1000 m long and 19 m
deep. The grid consisted of 10 horizontal elements (100 m each)
and 38 vertical elements (0.5 m each), for a total of 429 nodes and
380 elements. The saturated depth of the aquifer was 7.5 m at the
right boundary and 4 m at the left boundary. This represented a
water table slope of 0.0034 m/m, the prevailing hydraulic gradient
in the  aquifer.
  The  physical input parameters in this unsaturated simulation are
identical to those  used in the saturated simulation and are shown
in Table 1.
  Boundary and initial conditions for the unsaturated simulation
were the same as  those for saturated  flow with the exception of
the vertical constant head boundaries. The total pressure at the
nodes  representing the water table along the vertical boundaries
was specified to be zero. The nodes beneath these were assigned
pressures equal to the hydrostatic pressure as was done in the
saturated simulation. The portions of the vertical boundaries in
the unsaturated zone were specified as no-flow boundaries.
  This simulation was conducted for a 10-year period, with time
steps of 1 day.

RESULTS AND  DISCUSSIONS
  The  primary output from a SUTRA simulation was the con-
taminant concentration at each node in the system and a plot of
relative concentrations at each time step. The relative concentration
was defined as the ratio of the predicted concentration at a point
to the  source concentration.
  The  simulations discussed here represent  two very different
groundwater systems.  At the North Hollywood Dump, the satu-
rated zone is relatively close to the surface (at  a depth of 3 to
4 m) and the contaminant of concern, chlordane, is moderately
soluble (1.85 mg/1) and biodegradable (half- life  = 9.6 years). At
the Hardeman County site, the depth to groundwater is much
greater (approximately 15 m) and  the pollutant, carbon tetra-
chloride,  is  very  soluble  (800  mg/1)  and  virtually non-
biodegradable.

Results of Groundwater Modeling at North Hollywood Dump
  The  concentration data were printed out every 10 days for the
first 100 days of simulation; every 50 days from  100 days to one
year of simulation; and twice per year thereafter.
  The  concentration  profile for the  saturated  flow case  at  a
simulation time of 21  days is shown in Figure l(a). The profiles
from saturated and unsaturated applications are shown in the same
                                                                              CONTAMINATED AQUIFER CONTROL    233

-------
figure for comparative purposes. By this time, leachate from the
waste has begun to enter groundwater. The area corresponding to
the waste disposal region is shown at the  top of the diagram.
  The concentration profile for the unsaturated flow case at a simu-
lation time of 21 days is shown in Figure l(b).  The contaminant
migrates vertically to the groundwater table and beyond, and moves
considerably slower than in the totally saturated case. The con-
centrations in the upper regions directly beneath the site are greater
than those in saturated flow due to the lack of dilution and slow
migration of the  solute. The approximate location of the  water
table is shown  on this figure.
                                DISPOSAL SITE
                                                                                                  DISPOSAL SITE
                   DISTANCE X FROM SYSTEM BOUNDARY. (•.)
                   (•) 8ซturซtปd Cป<
                              DISPOSAL SITE
                              C7C0 . 0.001
                   DISTANCE X FROM SYSTEM BOUNDARY. (• )
                   (k) Urmlur.l.< Caป

                           Figure 1
          Relative concentration (C/C^ profile at l = days,
           Hollywood Dump (Note vertical exaggeration).

  The concentration profile for the saturated flow case at a simu-
lation time of one year is shown in Figure 2(a). The contaminant
plume is being affected by both vertical infiltration and horizontal
groundwater flow by this time.  Nearly vertical isoconcentration
lines are observed in the far field of the aquifer. Beneath the waste
site the plume has begun  to spread in the direction of groundwater
flow.
  The concentration profile for the unsaturated flow case at a simu-
lation time of one year is shown  in Figure 2(b). Note that as soon
as the plume reaches the water table and the saturated zone it
spreads in the direction of groundwater flow to create nearly vertical
isoconcentration lines in this saturated zone. Above the water table,
the contaminants continue to migrate very slowly  in the vertical
direction. The concentrations directly beneath the site are approxi-
mately 35% greater than those observed in the saturated simula-
tion (0.45 for unsaturated and 0.33  for saturated).
  The concentration profile for  the  saturated flow  case at a
simulation time of 7 years is shown in Figure 3(a). The profile is
                                                                                      DISTANCI * FROM SVSTIM BOUNDARY. {•.)
                                                                                      (•) Sstvrat** €•••
                                                                                                   DISPOSAL SITE
                                                                                                             -r
                                                                                        too        no        100         o

                                                                                      DISTANCE I FROM SYSTEM BOUNDARY. <•>.)
                                                                                              Figure 2
                                                                            Relative concentration (C/C^) profile ai t = I year,
                                                                              Hollywood Dump (Nole venicaJ exaggeration).
very similar to that obtained at 1  year of simulation time except
that concentrations have reached  steady-state values by 7 years.
The steady-state relative concentration in the far field was found
to be approximately 0.30.
  The concentration profile for the unsaturated flow  case at a
simulation time of 7 years is shown in Figure 3(b). The system had
reached steady-state by this time and the relative concentration in
the far field had reached a value of 0.275.
  Figure 4 shows the relative concentration versus time relation-
ship at the farthest point from the  waste site (x =  400 m) for both
saturated and unsaturated conditions. To compare the results of
the two simulations, the steady-state concentrations and break-
through  times for various concentrations were compared.
  The factor of safety in assuming  that the region beneath the waste
was totally saturated was defined as the ratio of the steady-state
concentration for the saturated case to the steady-state concentra-
tion for the unsaturated case. Both of these values were the maxi-
mum concentrations  measured at  x  = 400 m.  For the North
Hollywood  Dump the factor of safety is

  FS =  0.304/0.277  =1.1

  The breakthrough time for a given concentration was defined
as the time for the point in the far-field (x = 400 m) to reach that
concentration value. The breakthrough times for various relative
concentrations  are shown in Table 2. (This table shows data for
both sites.) The time lag for the contaminant plume in unsaturated
simulation ranges from 0.09 years at C/Co = 0.001 to 3.1  years
at C/Co =  0.275.
 234    CONTAMINATED AQUIFER CONTROL

-------
                                DISPOSAL SITE
      o


     1.4


     2.8


     4.2


     5.6


     r.o


     6.6
                   Table 2
   Breakthrough Times for SUTRA Simulation
                    DISTANCE X FROM SYSTEM BOUNDARY. (HI.)
                    (a) Saturated Caaa
                                DISPOSAL SITE

                    DISTANCE X FROM SYSTEM BOUNDARY, (m.)
                    (b) Unaatvratad Caaa
                           Figure 3
        Relative concentration (C/C0) profile at t = 7 years,
          Hollywood Dump (Note vertical exaggeration).
    0.3
   0.26
 O
O
   0.20
IU 0.15
O
z
o
o
111
2 0.10
UJ
cc
   0.05
   0.00
                                         • Saturated
                                         • Unaaturatexl
                  2468          10

                           TIME, years
                           Figure 4
   Relative concentration vs. time at x=400 m, Hollywood Dump
North Hollywood Dump
C/C Breakthrough Times (yr) Time
Saturated Unsaturated Lag







.001 0.19
.01 0.41
.10 1.10
.20 1.70
.25 2.30
.275 2.90
.30 4.50
0.28 0.09
0.55 0.14
1.50 0.48
2.40 0.70
3.50 1.20
6.00 3.10
	 	
Hardeman County Site










C/C Breakthrough Ti
Saturated Dns
.001 0.19
.01 0.40
.10 0.90
.15 1.10
.18 1.30
.20 1.50
.25 2.20
.268 3.50
mes (yr) Time
aturated Lag
0.69 0.50
1.10 0.70
2.60 1.70
4.00 2.90
7.00 5.70
	 	
	 	
	 	
DISPOSAL SITE
0
I '
i

a.
o
3

4



' ' ( '
) I
_ tt ป
0 0 ซ
o o o
o o o
-

V V
1000 800 600 400
DISTANCE X FROM SYSTEM
(a) Saluratad Caaa
v^oa^y -
^ 0.03 _/
/

/
M I
O/
dl
/
200 0
BOUNDARY, (m.)

                             DISPOSAL SITE
 1000     600      600     400      200       (
        DISTANCE X FROM SYSTEM BOUNDARY, (m.)
        (b) Unaaturatad Caaa

                   Figure 5
Relative concta.tration (C/C^ profile at t = 21 days,
Hardeman County Site (Note vertical exaggeration).
                                                                                  CONTAMINATED AQUIFER CONTROL    235

-------
                                    DISPOSAL SITE
                                                                                                         DISPOSAL
                 _^
                                        ~r
                         _L
                              V   V.    /
                                   400
                                           200
                  100      100

                DISTANCE X FROM SYSTEM BOUNDARY. (m )
                (•) Sitvr alcd Gซ ••
                                    DISPOSAL SITE
                DISTANCE > FROU SYSTEM BOUNDARY. (ป.)
                (k> UnปUral*ซ Cซtซ


                           Figure 6
          Relative concentration (C/C^ profile at t= 1 year,
         Hardeman County Site (Note vertical exaggeration).
          1000     100     ซ00      400     200       0

                 DISTANCE > FROM SYSTEM BOUNDARY. <- )
                                                                                                        DISPOSAL SITE
                                                                                                       4OO
                                                                                                               ZOO
          1000     *00     100
                 DISTANCE X FROM SYSTEM BOUNDARY  (- )
                 (B) Uปiซlซrซl*ซ €•••


                           Figure 1
        Relative concentration (C/Ca) profile at t= 10  years,
        Hardernan County Site (Note vertical exaggeration).
Results of Groundwater Modeling at Hardernan County Site
  The output for the Hardeman County disposal site was collected
every 10 days for the first 100 days of simulation;  every 50 days
from 100 days to 1 year of simulation; and twice per year thereafter.
  The concentration profile for the saturated  flow condition at
a simulation time  of 21 days is shown in Figure 5(a). By this time,
leachate from the  waste had begun to enter groundwater. Because
of the shallow nature of the aquifer, the vertical variation in con-
centration is small even at  early times because there is  nearly
instantaneous mixing in the system.
  The concentration profile for the unsaturated case at a simula-
tion time of 21 days is  shown in Figure 5(b). The approximate
position of the water table also  is  shown  on the figure. At that
time, all of the solute  was above the water table and moving
vertically toward  the saturated zone.
  The concentration profile for the saturated case at a simulation
time of 1 year is shown in Figure 6(a). The isoconcentration lines
at that time are all nearly vertical, and the relative  concentration
in the far-field  (x = 1000 m) is 0.118.
  The concentration profile for the unsaturated case at a simulation
time of 1 year is shown in Figure 6(b). The difference in the satu-
rated and unsaturated zones is very evident in this profile. Above
the water table the isoconcentration lines are horizontal as the solute
slowly moves downward to the water table. When the contaminant
reaches the  saturated zone the concentration contours become
almost vertical because of the nearly instantaneous mixing in the
aquifer. The far-field relative concentration (x  = 1000 m) had
reached a value of 0.007 by  that time.
   The concentration profile for the saturated case at steady-state
(after 5 years) is  shown in Figure 7(a). As with the contaminant
plume after 1 year, the vertical change in concentration at any point
in the system is minimal. The relative concentration at x  = 1000
m had reached a value  0.268 by this time.
   The concentration profile for the unsaturated case at steady-state
(10 years) is shown in Figure 7(b). The profile is very similar to
the one that had developed after one year, but concentrations were
higher. The far-field (x = 1000m) relative concentration was 0.180
at this time.
  The relative concentration versus time relationship for saturated
and partially saturated flows is shown in Figure 8. These concen-
trations were observed at the boundary 1000 m from the fixed con-
centration  boundary upgradient of the site.
      o
      I 010
      O
      U O.K
      O
      o
      u
                            TIME, year*

                            Figure 8
        Relative concentration vs. time, Hardeman County Site,
                          at x = 1000 m.
 236    CONTAMINATED AQUIFER CONTROL

-------
  The factor of safety obtained by assuming a totally saturated
system is the ratio of the steady-state saturated concentration to
the steady-state unsaturated concentrations (at x = 1000 m from
the fixed concentration boundary). For the Hardeman County
Disposal Site, this factor of safety is

FS  = 0.268/0.180  = 1.5

  The breakthrough times for various concentrations are shown
in Table 2. The  concentrations are those that are measured at a
distance of 1000  m from the upgradient fixed concentration
boundary. The time lag for the contaminant plume in the unsatu-
rated simulation ranges from 0.50 years at C/Co  = 0.001 to 5.7
years at C/Co = 0.18. These time lags are considerably greater
than those for the North Hollywood Dump  site due to the greater
distance to the saturated zone at the Hardeman County disposal
facility.

CONCLUSIONS
  The objective  of this study was to assess the merit of including
the unsaturated zone when modeling the subsurface flow and trans-
port of contaminants from a hazardous waste disposal facility. To
model unsaturated flow, it is necessary to obtain a great deal more
field  data than  are required to model the system as a totally
saturated aquifer. In addition, the computer time necessary to run
an unsaturated simulation using the groundwater model SUTRA
was considerably greater than that required for a saturated simu-
lation because of the non-linear nature of unsaturated flow equa-
tions and the iterative solution techniques  used to solve them.
  The factor of safety as defined in the report was less than two
for both waste sites considered in this study, (i.e., the saturated
concentration was less than twice the unsaturated concentration).
The uncertainties that are involved in parameter estimation for the
system may cause variances this great, or greater. It is the authors'
opinion that the inclusion of the unsaturated zone probably is not
warranted in cases where the disposal facility is located in humid
regions (i.e., shallow water table).
  This comparison may not be representative of arid regions where
the depth to the water table is considerably greater than either of
the two cases considered in this report. The authors recommend
that if further study of this topic is pursued, arid region disposal
sites should  be included so that this question  might also  be
answered.


ACKNOWLEDGEMENTS
  This study was performed as  part of the primary author's
Master's research at Vanderbilt University and was funded under
the Department of Energy's Nuclear Waste Management  Intern-
ship Program.


REFERENCES
1.  Voss, C.  I. "SUTRA—Saturated—Unsaturated Transport." United
   States Geological Survey; Reston,  VA, 1984.
2.  Moore, J. E. "Contribution of Groundwater Modeling to Planning."
   Journal of Hydrology 43, 1979, 121-128.
3.  Broshears, R. E., "A Conservative, Probabilistic Risk Algorithm for
   Hazardous Waste Sites." Ph.D. dissertation, Vanderbilt University,
   Nashville, TN, 1979.
4.  Bear, J.  1972. "Dynamics of Fluids in Porous Media. American
   Elsevier,  New York, NY, 1972."
5.  Wang, H. F. and M.  P. Anderson. "Introduction  to Groundwater
   Modeling." W. H. Freeman and Company, San Francisco, CA, 1982.
6.  van Genuchten, M. Th. "A Closed-form Equation for Predicting the
   Hydraulic Conductivity of Unsaturated Soils." So/7 Sci Soc ofAmer
   J. 44:  1980, 892-898.
7.  E. C. Jordan, Inc. "North Hollywood Dump Environmental Assess-
   ment and Action Plan.", 1983.
8.  ERM-Southeast, Inc. "Environmental Evaluation and Assessment of
   Control Measures at the Velsicol Disposal Site." Brentwood, TN, 1985.
9.  "Leachate Migration from a Pesticides Disposal Site in Hardeman
   County, Tennessee, United States Geological Survey, Nashville, TN,
   1978.
                                                                                CONTAMINATED AQUIFER CONTROL    237

-------
                           Remedial Action  Evaluation  System:
                 A Methodology  for  Determining  the  Completion
                        Point  for Aquifer  Restoration  Programs
                                               Paris Hajali, Ph.D.
                                         Woodward-Clyde Consultants
                                              Santa Ana, California
INTRODUCTION
  The primary objective of this study is to develop a methodology
to evaluate remedial actions at uncontrolled hazardous substance
sites. The literature includes a small number of systematic evalua-
tion  procedures  for selecting the  types  of cleanup methods;
however, the need exists for a methodology to evaluate remedial
actions in terms of their degree of cleanup.
  Three different tasks were accomplished in preparation for the
development of the methodology and they are: (1) review of the
literature; (2) conduction of a national survey regarding aquifer
cleanup issues; and (3) review  of 45 remedial action case studies.
Many findings were obtained, and documented and used to develop
and structure the  methodology. The developed methodology is
called the  "Remedial  Action Evaluation  System" ("RAES",
pronounced as race).
  RAES attempts to evaluate the extent of cleanup based on site-
specific information and resource use after implementation of a
remedial action. Site-specific information and water use are the
most desired and  practical  bases for planning aquifer remedial
actions as suggested by the literature and indicated by the survey.
In addition, reviewed remedial actions indicated that site specific
information and existing groundwater use greatly affected the selec-
tion of cleanup measures and techniques.
  At several points in the methodology, preferences or values were
attributed to various entities for quantification  purposes. These
preferences and values were assigned based on the findings and
suggestions of the preparatory tasks.

METHODOLOGY: STRUCTURE AND USE
  This section describes the Remedial Action Evaluation System
to be used in evaluating the extent of cleanup at uncontrolled
hazardous waste facilities. Detailed instructions for using RAES
are provided in the following  sections. RAES is a methodology
that  requires uniform  and consistent professional  judgment
regarding its application. It does not address the degree of site
cleanup in terms of safe or acceptable concentration levels. Rather,
the degree of cleanup is approached in terms of public health and
environmental protection, effectiveness of remedial action tech-
nologies, resource use and potential contamination.
  RAES distinguishes between the on-site and off-site regions to
simplify and organize the evaluation process. It is important to
identify the regions along with their corresponding cleanup efforts.
For example, the effectiveness of an on-site slurry wall may be
different from the one of an off-site slurry wall. Therefore, it is
suggested that the boundaries  of the on-site and off-site regions
be determined prior to RAES utilization. On'site is defined in this

238    CONTAMINATED AQUIFER CONTROL
study as the region of the facility or the source of contamination
and off-site is defined as any affected region outside of the on-site
region.
  In addition, the RAES is structured to evaluate migration routes
separately: air,  surface soil, unsaturated zone, groundwater and
surface water. In cases where a route is not affected or not within
the on-site and  off-site boundaries, this route is eliminated from
the evaluation. The dissection of the site and migration routes serves
as a simplification means to ensure the inclusion of all involved
factors and to provide a systematic evaluation procedure. RAES
procedures are  graphically  illustrated in Figure 1.
                           Figure 1
                   RAES Flow Chart Procedure
  RAES assigns four score indices to remedial  actions at  a
contaminated site, termed RA models, and they are defined as
follows:

-------
• RA Objectives (I0)—reflects the degree of remedial action con-
  tribution to the protection of public health from hazardous sub-
  stances
• RA Technologies (IT) reflects the degree of short and long term
  effectiveness of  the implemented remedial action
• RA Potential  Contamination (IpC)—reflects the potential  for
  harm to  humans or  the environment  from migration of
  hazardous substances away from the remained contamination
  by routes involving air, surface soil/water, unsaturated soil, or
  groundwater
• RA Resource  Status (IRS)—reflects the possible resource  use
  (water and land) after the remedial action in contrast to the use
  before the contamination.

  The score index for each RA model is obtained by using  the
procedures explained in the following sections. Each procedure con-
siders a set of pertinent factors regarding the site and remedial ac-
tion, which is quantitatively evaluated to yield a score index. Score
indices are then combined as follows to obtain the final score in-
dex for the implemented  Remedial  Action  (IRA) under con-
sideration:
          1IRA
                             (i)
  The equation used to calculate ^ has a deemphasizing effect
on the scores to give more consideration to the smallest value(s).
Therefore, IIRA is the composite score index for the four models
that gives a numerical indication of the degree of cleanup.
  Used by itself, the  composite index (IIRA) does  not lead to a
meaningful conclusion. However, obtaining other indices on the
same scale would allow some comparison that leads to some con-
clusions.  Furthermore, it  would  be extremely meaningful  to
compare the implemented remedial  action score index (IIRA) to the
score index of the no action alternative. This would allow the user
to quantitatively show the site condition improvements in terms
of the score indices. The score index for the no action alternative,
INA, can be obtained by using the RAES and selecting the worst
possible case for the site under consideration. INA is calculated in
a similar  fashion as IIRA.
  Another useful score is the index representing the best possible
remedial action regardless of cost. The best possible remedial action
(BRA) relates to the best and most effective technology available
to prevent, reduce or eliminate contamination within an accept-
able period of time. This score index, IBRA, may be obtained by
considering the best possible measures and technologies in the
evaluation. The IBRA calculation is also similar to IIRA.
  Once  the  score indices  for the:  (1) no  action  (INA), (2)
implemented remedial action (IIRA) and (3) best possible remedial
action (IBRA) are obtained, their relative position on the same scale
would be similar to Figure 2. The interpretation of the score indices
will be discussed later in this section.
           Accoipllihid
           laprovnnt
      IHA
. Foillblo
    lapromcnt
__ frlitlo*
  Condition!
                                          IIIA
Ho Action

   or
Worst Cซiซ
                       45
                     Xraปdi>l
                     Action
              lซt poiilkU
                 lil Action
                      of co*tซ)
                           Figure 2
       Relative Comparison of Remedial Action Score Indices
                     Remedial action costs play a major role in the selection decision
                   process. In fact, the NCP requires the U.S. EPA to select the lowest
                   cost alternative that effectively mitigates and minimizes damages
                   and provides adequate protection  of public health  and  the
                   environment.
                     RAES addresses cost of cleanup actions by first calculating or
                   estimating the cost of the best possible remedial action attained
                   through the evaluation  and then comparing it to the cost of the
                   implemented remedial action. This cost comparison combined with
                   the score indices comparison will give a better and more realistic
                   picture of the differences between  the  implemented  and best
                   possible remedial  actions.

                   INTERPRETATION OF RAES RESULTS
                     Following the procedure discussed above, three score indices
                   would be obtained: (1)  score index for the no action alternative
                   or the  worst case (INA),  (2)  score index for the implemented
                   remedial action (IIRA) and (3) score index for the best possible
                   remedial action (IBRA). To interpret the  results, the following steps
                   must be accomplished:

                   • The difference between INA and IBRA will be considered to be
                     100%,  meaning the IBRA is the highest possible level  achiev-
                                                                     ble at the present time.
                                                                     The difference between I
                                                                  NA
and
                                                       and
                                                                                                  be
                                           expressed in percent based on the previous step.
                                         • The costs for the implemented remedial action and the best pos—
                                           sible remedial action will be compared and expressed either in
                                           a percentage or dollars format.

                                           These differences may be stated as a percentage of improvement
                                         from the no action reference point and as a percentage of possible
                                         improvement coupled with cost figures. It should be noted that
                                         improvement is defined, in this study, as score index improvement
                                         since the relationship of actual improvement to the score index
                                         improvement is not established. The interpretation of such findings
                                         may vary from one site to another depending on the type and extent
                                         of contamination and financial situation of the  involved parties.
                                           RAES could also  be used  as a tool to show the relationship
                                         between improvement and cost. By obtaining several score indices
                                         (e.g., at  10 or 20%  intervals) between INA and IBRA along  with
                                         their corresponding costs and plotting them on an x-y graph, a curve
                                            Colt
                                            ($)
                              IMA

                           Mo Action
                     'BRA

                     Beit Fouible
                     Reaedial
                     Action
                                                                    Figure 3
                                                       Hypothetical Curve of Remedial Action
                                                             Improvement Versus Cost
                                                                               CONTAMINATED AQUIFER CONTROL    239

-------
would be formed displaying the rate of improvement versus the
increase in cost (see Figure 3). Therefore, RAES may be used as
a tool to find the most cost-effective alternative or remedial ac-
tion.  It is possible to have RAES and the unit costs of remedial
action measures and techniques incorporated into a computer pro-
gram to allow a faster and more efficient evaluation.
  RAES could also be used to compare remedial actions at different
sites.  This may be justified  by stating that RAES considers site-
specific conditions and allots weights or  importance  to various
factors in order to translate the true situation of a site into the
methodology. Since professional judgment is involved in utilizing
RAES, it would be advisable to have one individual or group apply
the methodology  to preserve the consistency in the evaluation.

REMEDIAL ACTION MODELS
  In this section the four models (1) RA Objectives, (2) RA Tech-
nologies, (3) RA Potential Contamination and (4) RA Resource
Status are discussed in detail. The discussion includes reasons for
including each model and a complete explanation of the quantita-
tive evaluation.

RA Objectives—10
  The NCP requires the U.S. EPA to select the lowest cost alter-
native that  effectively mitigates and  minimizes damages and
provides adequate protection of public health, welfare and the
environment. To meet such a requirement, the selected remedial
action has to include measures that specifically deal with the con-
taminants and their migration routes.
  In this part of  the methodology, the objectives are evaluated
according to the measures and technologies included in the remedial
action in terms of protecting the public health and the environ-
ment. The effectiveness and costs are considered in other sections.
  The evaluation  of the RA objectives is  based on the migration
and damage reduction measures included in the implemented
remedial action. To completely address this issue, aU migration
and consumption modes must be considered. The contaminant
routes to humans through the various migration and consumption
routes are illustrated in Figure 4. Therefore, to protect the public
health, the consumption modes (inhalation, ingestion  and direct
contact) must be eliminated or reduced to safe levels. Two alter-
natives may be used to achieve such a task, including relocation
of the affected population or stopping/reducing the migration of
contaminants.  Since the former alternative is  not a cleanup
measure, the evaluation will only consider the latter. Three dif-
ferent measures may be taken separately or combined to stop or
reduce migration  of contaminants  from a polluted site. These
measures consist of containment, treatment and removal of con-
taminants or contaminated material.
                NlfMtlo*                 Con.uaptlon


Coatmiiuntl



Alr/Atw>ปptMr*

loll
toot Ck.U Unlซ4l/ri.nt.>

W.It.
toll
U.t.r (Ioerปttoiul)
—





1 oh. 1. lion




Direct Contact






                          Figure 4
           Contaminants Route to Humans—Migration
                   and Consumption Modes
  The development of the quantitative evaluation for the RA ob-
jectives was based on the concept of how  much  the  remedial
measures contribute to the protection of public health, which is
considered to be the first degree objective. As stated earlier, the

240    CONTAMINATED AQUIFER CONTROL
 evaluation was tailored  to  deal  separately  with  the various
 systems/zones  for the on-site and off-site contamination. The
 procedure and the instructions for the evaluation are shown in
 Table 1. The procedure is shown graphically in Figure 5 and can
 be summarized as follows:

 • Identify systems (or migration modes) involved, then assign an
  importance value to each involved system according to its con-
  tribution to disperse contamination or affect human health. The
  sum of all  importance values should always equals 10 regard-
  less of the number of systems involved. The system importance
  values in the evaluation is included to  reflect a more accurate
  picture of  site conditions.
 • Identify actions/measures included in  the  remedial action to
  eliminate/prevent/minimize contamination impacts.
 • Evaluate the identified actions/measures in terms of their con-
  tributions to the first degree objective  by system/zone for the
  on-site and off-site regions. The contribution to the first degree
  objectives are categorized and assigned weight values according
  to their achievement. The categories and their corresponding
  assigned values are summarized in Table 2.
 • Tabulate and sum the assigned values by system regions as shown
  in Table 1. The result is the RA objective score index I0, which
  will be a part of the final score index, I1RA.
•ad
•cify iavolv
•••if* tซfor
tซBCซ TallM
1
litatltj tctiamd
prevent /•iBlaifr
•IfTitiM of co
) ซyl07ซ4 to
•/•iUlMM
1
Doซ* efc* acciซซ protect
ttM pซbiic IU*IUT
                                                                                             Figure 5
                                                                                 R.A. Objective Evaluation Procedure
  The maximum value for I0 is 100 and the lowest is 0, where 100
indicates that remedial action objectives, which are the protection
of human health, were met. Various remedial action technologies
with their corresponding objectives and functions are included in
Tables 3 and 4.  These tables may be regarded as  a directory or
an aid to help identify the objective(s) of remedial measures.

RA Technologies—IT
  The extent of cleanup at a contaminated site is a function of
the employed remedial action technologies. Remedial actions con-
sist  of four basic categories: (1) no action,  (2) containment, (3)
treatment, and (4) removal. Among the four categories, removal
of contamination is  the  most effective  action  (not necessarily
feasible) and the action is the least effective.

-------
                          Table 1
       Work Sheet for R.A. Objective Score Index - IQ
                 Conaupption Mode*
                           Direct
                lition  Inflation Contact
levortonc
 Value
  (A)
                                                        SCOT*
                                            t.A. Objective    Indea
                                              A*ai|iied     by
                                              Value       Syateu
                                                      (C)
Air

Surface Soil

Unaeturated flail

Ground Hater

Surface Water
l.A. Objective Sear* Indซ  HO,
N.A.t  Mot applicable
  Xi  Applicable

*Croun4 vater releaaea to aurface vatar (lake, river*) •prlng, etc.)

Calculation Proojjarei

(A) Value*  for  tbie  coluea  ere eaaitoed  according  to (7*ten contribution to  dUperae
   contamination or affect hman* (Step 1 in the  text), the  auimation of all value* in (A)
   ehould equal 10 at all tlปa*.
<•) Value* for thia colwn are obtained froei Table T.3.
(C) Value* for thia coluen are obtained by Multiplying value* in  (A) end (B) ซC) • (A) * (B)).
R.A. Objective Score index IQ i* the meMation of value* in eoluan (C).
                            Table 2
         Assigned Values for Remedial Action Objectives
lat Degree Objective  ia

Aceompliahed

Partially Accomplished

  High
  Medium
  Low

Mot Aceompliahed
                             10
Accomplished:
                      means that the  implemented action  eliminates
                      impacta  on  humans in  terms  of  inhalation,
                      ingeation, and/or direct contact.
Partially Accomplished:  mean!  that  the  implemented  action doei not
                      eliminate  impซctซ  on  humana)  however,  it'
                      reduce!  the  impacta.     The  reduction  of
                      impacta  ia  categorised into  three  claaaea
                      (high,  medium, and  low)  which reflect the
                      extent of human impact reduction.
Rot Accompliahed t        meana  that  the  implemented  action doea  not
                      eliminate nor reduce the impact on humani; or
                      no remedial action hai  been employed.
                            Table 3
      Objectives/Functions of Remedial Action Technologies
                                     11
                                                 ill
                                                 !::
                     iii    j,fli    ij]    mi    in    mm
 !••ป>• VlKt ซ*MIU|
                   111  i     ii
  In  each  remedial  action  category,  there  is  one  or  more
measure/technology,  except for no action.  These measures and
technologies vary in application and effectiveness regarding limita-
tions  and levels of achievement. In addition, the measures/tech-
nologies differ in terms of development. Some techniques are more
developed, meaning well proven, therefore, information on long-
term effectiveness is available. Other measures are less developed,
and therefore, long-term information is not readily available.
                                                              Table 4
                                               Objectives of Remedial Action Techniques

                                     Based on the above discussion, the evaluation of remedial action
                                  technologies was included to reflect the degree of short-and long-
                                  term effectiveness of the implemented remedial  action.
                                     To quantitatively evaluate remedial actions, values were assigned
                                  to RA categories and technologies. The scores were then set in an
                                  organized manner to be combined for a selected remedial action.
                                  The result is  a score index for the RA  technologies. Again, the
                                  evaluation is done by system/zone considering the on-site and off-
                                  site regions (Tables  5 and 6).
                                     The four remedial action categories (no action, containment,
                                  treatment and removal)  were assigned values according to their
                                  effectiveness  to  cleanup a contaminated site  and to achieve  a
                                  contaminant-free  environment. The assigned values  for each
                                  category are summarized in Table 7. The effectiveness of the four
                                  categories is also a function of the system involved. For example,
                                  while capping prevents or reduces infiltration and contaminant
                                  percolation from the unsaturated zone, the horizontal flow of con-
                                  taminants in the  saturated zone is not  prevented. Therefore,
                                  capping is more effective toward the unsaturated zone than toward
                                  the groundwater. Table 5 includes the remedial action values by
                                  system. No action means that measures have not been taken to
                                  prevent, reduce  or eliminate contamination. Monitoring is cate-
                                  gorized as a no action measure. However,  it is considered a better
                                  measure than doing  nothing at all. Containment includes surface
                                  and subsurface controls such as caps, slurry walls and hydraulic
                                  wells. Treatment encompasses biological, chemical and physical
                                  treatment. Treatment is generally any measure that attempts to
                                  reduce the concentration levels of contaminants. Removal is con-
                                  sidered to be any measure that physically removes contaminants
                                  and/or contaminated material from their initial location to either
                                  on-site or off-site disposal facility.
                                     Remedial action technologies are assigned values based on then-
                                  applicability, effectiveness and confidence. Applicability refers to
                                  the extent of dependency of the technology on either the site charac-
                                  teristics or the waste characteristics. Therefore,  a technology of
                                  little or no site dependency is  considered  "very broadly applica-
                                  ble" and is assigned a higher value than a technology that is limited
                                  to sites of specific characteristics. Effectiveness refers to the extent
                                  of elimination, prevention or treatment achievable by the  tech-
                                  nology. A technology that can produce a "leak tight" contain-
                                  ment or high level of treatment is assigned a higher value than a

                                              CONTAMINATED AQUIFER CONTROL     241

-------
                           Table 5
         Work Sheet for R.A. Technologies Total Scores
                              Table 6
            R.A. Technologies Score Index - IT Work Sheet
Sytcin
Air
Surface Soil
Cround Water



lyeteti
Total Score Importance
(Table ป.*) Value
(A) (1)



Z(C>
IT
Score
Index
•y lyeta*
(C)




*&""
                                                                       Celculetioo frocซdureป:

                                                                       (A) ii obtained (roป Table V.6

                                                                       (B) if  the  ayateii  iatportaace   value  aaaia^ad  according  Co  vyiten
                                                                           contribution to dieperoe  conteeiinatiow or  affect  hyaena  (aaw at
                                                                           coluK (A)  in  Table V.2).  The tuanutkon of valttae U col too. (S)
                                                                           thould be equal to 10 ml all  tiitat.

                                                                       (C)  it (A) • (B)

                                                                       i.A.  Technolositt Score Indem - IT •  ^^ -  100

                                                                       Where 30,000 it Che .ktxiauei pottible ปaluซ for  HO.
 Kem - (ป) • (|) • (C) • (ป)


 -Wn (I) i* th* •.A. *•.•ป U cBlซ.n D.I.F.C, n 1.


. II taWT* IhkKB OM tซCm.h.>l.>Cr tM* •>•*ป ••.*)Ctซ*> ซปซlซf tlM ••••


 lc*m - I'**, * ซ.I U)j *... * 0.!(*)•,) • (*) * (ซ • (I)


Mr* I nrr*ป*.t* I.JL.  *il-ซt t* colva.-.* C,D,t,r, M 0| (A)
                                 fซr (
     Tl* valM U €•!••• (C) U • |MC*T nf
     •••t(M4 by tkซ ซntMt*r.  (C) *•• ซ
     F*r au^l*. If •ซ!•<• CซFซ *•*• ปซ•
                              i UIMI ซr r*fjปIUl MUM. nil fttt*r i.
                              I.M of It (ซr • c*^lซtซ vw*elUI  •ctl.->.
                              •VM Cte WI..1* CMtซBปMt*>*) ซttป, tk>t* I 1.
                     • _.!_ ..!ซ .1 KOD. WOC U >••ซ - ck>
                     !ป•  CMCMiMBt* •* CซซIMlutซ4 WtnUl.  tfT
     tb. )000 ซtaU t. ซt.rw .. MOO.

  *.  Tปtat •cart  t. tbซ MVMILM *f itl Mffni f*r Mซk irctM.

  }.  OM ••ป• •rcajlM LI crfll**! ซlMr*lซซ ••! •pvUuปla.
technology that allows some leakage or cannot produce high quality
effluent. Applicability and effectiveness were combined to obtain
a technology "effectiveness" value. Table 5 includes the assigned
values of various technologies. Confidence relates to the develop-
ment stage of the technology in terms of data availability on long-
term effectiveness and experience. Therefore, a well proven tech-
nology with readily available data on long-term effectiveness will
be assigned a higher value than a technology with no data availa-
ble or with limited experience. Table 5 contains the assigned values
for confidence in technologies.
  The assigned values for the various  factors are tabulated  in a
manner that when evaluating a selected remedial action, they are
easily combined to result in: (1) a  score for a single technology,
(2) a total score for the RA technologies by system, and (3) a score
index (IT). The tabulated values and evaluation format are illus-
trated in Tables 5 and 6.

RA Potential Contamination— IPC
   Since  most remedial actions do not  completely  remove all the
contaminants and contaminated materials from polluted sites  (on-
site and  off-site), potential contamination  is possible. To account
for such a possibility, RAES was developed to include  an evalua-
tion of potential contamination after  site cleanup.
                                                                                                  Table 7
                                                                                Assigned Values for Remedial Action Categories
                                                                         t.A.
                                                                       Category
                                                   Attigned
                                                    Score
 Ho Action

 Containment

 Treatment

 traovtl
 0

 3.5

 6.5

10.0
  The development of the potential  contamination  evaluation
procedure was founded on the following questions:

• Are there any contaminants remaining on-site and off-site? What
  are their characteristics?
• Are there any targets (population, water supply, etc.) that might
  be affected. What are they?
• What are the migration routes that might lead to potential harm
  to humans or the environment? What are their characteristics?
• What measures were taken (remedial action) to reduce, prevent or
  eliminate harm to humans?

The potential contamination is evaluated based on four primary
factors:  (1) proximity  to targets  (population, sensitive environ-
mental areas and resources), (2) hydrogeologic characteristics, (3)
contaminant characteristics and (4) implemented  remedial action.
These factors  reflect the potential for harm to humans  or the
environment from migration of hazardous substances from con-
taminated  areas  by routes involving  air, surface, soil/water,
unsaturated soil or groundwater  (Table 8).
  The Hazard  Ranking System (MRS) employed by the U.S. EPA
was used with minor modifications (in the calculation process and
other parameters shown in Table  8 and Figure 6) to assign values
to the various routes, targets and waste characteristics. The assigned
values to these factors are shown in the MRS tables.  These values
are assigned according to migration routes as indicated in Table 9.
242     CONTAMINATED AQUIFER CONTROL

-------
                         Table 8
           Contaminants Concentration Remained
                                                                                             Table 9
                                                                            Work Sheet for R.A. Potential Contamination
VC*(Z)
0
10
25
50
>50
Assigned Value
0
1
2
3
4
*VC -   (Ct _ Cg)/Ct   * 100
where Cf  is concentration of remained contaminant  at
time t  (t is time of  evaluation), and

Cs is known safe levels of contaminant in concern  (or
a desired established level).
If Cs  is not known or established  then the assigned
value  for concentration is 4.
  The score for each route is obtained by considering a set of
factors that characterize the potential of the conditions to cause
damages (Table 9). Each factor in Table 9 is assigned a numerical
value according to prescribed guidelines. This value is then multi-
plied by a weighting factor (termed multiplier in Table 9) yielding
the factor score. The factor scores are then combined as follows:
scores within a factor category (for example route characteristics)
are added (shown in Table 9 as A, B, C, and D); these scores are
then multiplied  together to result in a score for air, surface,
unsaturated zone and groundwater routes for each category.
  The RA potential contamination score index is then obtained
by summing all total scores and dividing the sum by 78,348 which
is the maximum possible score that could be obtained using the
work sheet in Table 9.  The score index is then expressed  as a
percentage.
                  RS
Resource Status—I
  In addition to human health impacts imposed by hazardous sub-
stances contamination, other negative impacts may be inflicted.
Such impacts may include cessation or reduction of agricultural,
industrial or commercial activities as  a result of air, land and/or
water contamination. Pollution does not physically reduce the
amount of resources available. Rather, it makes many uses more
expensive, difficult  and in some cases impossible. For example,
agricultural uses may be completely or partially stopped because
the irrigation water, after contamination, is not suitable for use.
Therefore, the evaluation of remedial actions should also include
an evaluation of the affected resources such as water and land use.
The evaluation in this case focuses on the change in use of resources
under consideration as a measure of remedial action efforts.
  Evaluation of the resource status  is based on the concept of
comparing the resource use after the remedial action to the use
before the occurrence of contamination (Table 10).
factor*
i. Air Route:

Prevailing wind direction
B. Target!
Land uซe
t. Renediil Action
Toxic icy
Concentration
E. Total Score (EA*EB*C*CD)
11. Surface Route:
A. Route Gharacteriatica
Facility • lope/ intervening terrain
1-year, 24-hour rainfall
Site accessibility
B. Tenet*
Surface water use
intake downstream
*.. Reaedi*! Action
Toxicity/Pereiatence
Concentration
t. Total Score aA*EB*C*CD)
III. Unaaturated Zone Routei

Net precipitation
Fer-eabilitr
Depth to ground water
v. ReaedUl Action

Concentration
E. Total Score (EA*EB*C*ED)
IV. Ground Water Routci
A. Route CharacterUtica
Aquifer gradient
B. Target a
Ground water usa
DUtance to neareat well/
population aerved
C. Remedial Action
u. Uaatc Characteriatica
Toxic Ity/Peraiatence
Concentration
Multiplier
3
I
2
1
1
3


1
2
1
3
2
1
1
1


1
1
2
1
1

1
2
3
L
1
1
1
Assigned
Value

EA
ฃB
c
CO


I A
EB
c
CD


EA
ฃB
c
CD


EA
CB
c
ED
Sum
of
Assigned Total
Value* Score


	 '

	 1


	 1
	 1
	 I
	 1


	 '
	 '

	 1



	 1

	 I
E. Total Score (EA*EB*C*CD>
eferenc*
igure V.T
able V.JO
able V.23
able V.I1
able V.2S
able V.29
able V.J*

able V.13
able V.14
able V.lS
able V.16
able V.22
able V.23
able V.2A
able V.2B
able V.29
able V.3A

Table V.I 7
Table V.lS
Table V.23
Table V.26
Table V.I 9
Table V.34

Table V.16
Table V.19
Table V.26
Table V.27
Table V.29
Table V.34

                                                                     •uxi.um poa>ible score for R.A. potential contamination, and* (fie aubitraction fro* 100 waa alow
                                                                     to >ake IpC conaUtent in direction with the other indices.
                                                                     Resources are categorized as water and land. Water consists of
                                                                   surface and groundwater which could be used for purposes such
                                                                   as drinking, agriculture, industrial, commercial or recreational
                                                                   Land as a resource reflects the use of land as a property for activi-

                                                                               CONTAMINATED AQUIFER CONTROL    243

-------
                     Prevailing Wind  Direction
                                                                                           Table 11
                                                                           Assigned Values for Surface/Ground Water KM
                                                                     Water. Uae
                                                                  Municipal

                                                                  Agricultural

                                                                  Potential1

                                                                  Industrial

                                                                  None
                                                                                                              Aaaigned Value
                                                    10

                                                     7

                                                     5

                                                     3

                                                     0
 ^Potential water use  me tat that  at preaent  water  it  not
  uied; however,  future uae  it poiaible.


                          Table  12
                 Assigned Values for Land I se
     Contaminated  Site
     Aligned Value

                          Figure 6
         Assigned Values for Prevailing Wind Directions
                          Table 10
                  Resource Status Work Sheet
                                                                                 Land Uae
                                            Aasigned  Value
                                                       - •ซ
                  I trmm Tablia V.lf a*4 ป.*0
                  I frซa Table* V.M • ** ป.J7
                  I I torn T.bi.. V.U a*ซ V.17
                  I frM TซH. *.)•
                  i la* obiatubla !• tha> sj*ซlw*tl**i.
ties such as agricultural, industrial, recreational, commercial or
other useful purposes.
  Water resource evaluation assesses the usefulness of water after
the completion of the remedial action and compares it to the use-
fulness of water before contamination.  Such a comparison pro-
vides an indirect indication of water quality changes. To quantify
the evaluation, water uses were assigned values according to use-
fulness and importance of water to society (Table 11). The assigned
values were attributed to water use based on the results of a na-
tional survey conducted for this study.
  Land as a resource is evaluated in  a  similar fashion to water
resources. Land evaluation indicates the change in land use as a
result of the contamination. Although land use is affected in many
cases by water quality, redundancy does not occur in the evalua-
tion since water use evaluation is an indication of water quality
as a resource, and land use evaluation is an indirect indication of
the actual use of water. The assigned values  to various land uses
are presented in Table 12. Assigned values decrease as land use
restrictions or limitations (due to contamination) increase.

244    CONTAMINATED AQUIFER CONTROL
 No restrictions (by  contaaination)

 Agricultural/Ccranercial

 Industrial

 Potential1

 Disposal  facility

 None
10

 7

 5

 3

 1

 0
                                                                   Potential  use  neans  that  at  present  land  cannot be
                                                                    uaed for any purpose;  however, future  uae is possible.
  For example, assume that land in the vicinity of a contaminated
site is used for agricultural purposes. After the remedial action,
the quality of water is not suitable for agricultural use. This results
in a  land use loss and may affect the community economically.
If the land prior to the contamination was not used for agricul-
ture  or any other use, then land use loss is not inflicted. However,
the loss is accounted for by the water use evaluation.
  The quality of water may not always be the factor to impose
limitations on the use of water. Limitation may be imposed due
to the  presence of a nearby contamination plume. For example,
pumping near a contamination plume  may adversely affect  the
extent and movement of the plume. Therefore, pumping may need
to be  stopped or  reduced to discourage  such adverse effects.
Lowering the pumping rate, even though the water quality is still
safe or acceptable, is considered an imposed restriction by the con-
tamination; if it is not addressed by the remedial action, then such
a loss must be accounted for. For this reason,  Table  13 includes
assigned values to reflect limitations and partial losses imposed by
contamination after the completion of the remedial action.
  The seriousness of a contamination emerges not only from the
extent and concentration of pollutants, but also from the com-
munity dependency on the contaminated resource. To account for
such a factor, importance values for water  and land use were
incorporated in the evaluation. For example, a community with
an alternate water supply would suffer  less  than  a community
without one. The assigned importance values for water and land
use are shown in Tables  14 and 15.

-------
                          Table 13
           Assigned Value Adjustment Due to Imposed
                 Limitation from Contamination
                         Table 15
          Dependency Importance Value for Land Use
Reduction
in Use
CO
0
10
20
30
40
50
60
70
80
90
100
Assigned Value
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
                         Table 14
          Dependency Importance Value for Water Use
              Source Dependency
                                               Assigned Value
  No alternate source:
    completely dependent oo contaminated source

  Limited alternate aource:
    moderately dependent on contaminated source
    minimally dependent on contaminated aource

  Available alternate aource:
    can depend on other aource
  Having explained the various factors involved in the resource
status evaluation, the procedure to calculate the resource status
score index (IRS) is presented in Table 10. The procedure simply
involves entering the appropriate assigned values in their proper
place in the table. The change in resource use is then obtained by
subtracting the values under the use after the remedial action from
the use before the contamination. The  value for the use after
remedial  action may involve  a reduction factor to account for
partial losses (as explained earlier). The score for the change in
                                                                          Land Use  Dependency*
                                            Assigned Value
                                                                   Community dependency on land  is:
                                                                      High
                                                                      Moderate
                                                                      Minimal
                                                                      None
                                                                   ^Dependency    could    be    economical,    agricultural,
                                                                    residential, etc.
use is then multiplied by the importance value to yield the final
resource use score. Finally, the final scores for groundwater, surface
water and land are summed to result in the resource status score
index (IRS).

CONCLUSION
  In summary, RAES was developed to evaluate remedial action
and, like any other methodology, it has some flexibilities and
limitations. To obtain good results, RAES must first be well under-
stood and, secondly, uniformly utilized. Furthermore, it should
be recognized that an ample amount of information is necessary
before the utilization of RAES. This means that a multiple dis-
ciplinary team participation is greatly desirable. Therefore, the
input of such a team to the methodology may prove to be helpful.
  RAES was developed  in a manner that  allows expansion or
further inclusions to its  structure. Such inclusions  may be the
emergence of a new technology for aquifer restoration or changes
in the effectiveness of a current treatment. It is, therefore, important
to update the information in the methodology as more data and
knowledge are introduced.
  The interpretation of RAES's results may vary from one site
to another depending on the type and extent of contamination,
the affected population and environment and the financial status
of the involved parties. This is due to the fact that RAES was de-
veloped to accommodate site-specific conditions. Such a flexibili-
ty gives RAES the capability of resulting in evaluation reflecting
the existing physical, financial and institutional site situation. The
intent here, however,  is not to claim that RAES is the ultimate
answer to the "how clean is clean" issue. Rather, RAES as a tool
should provide some indications regarding the degree of cleanup
achieved by remedial actions in terms of human protection, water
use and land use and not in terms of contaminants concentration
in water or soil. Therefore, in the absence of standards and guide-
lines, RAES is a useful method to quantitatively address the issue
of "how clean is clean."
                                                                                CONTAMINATED AQUIFER CONTROL     245

-------
                      Paper withdrawn by author
               While printing of Proceedings in progress.
246  CONTAMINATED AQUIFER CONTROL

-------
       Paper withdrawn by author
While printing of Proceedings in progress.
                            CONTAMINATED AQUIFER CONTROL  247

-------
                    Paper withdrawn by author
             While printing of Proceedings in progress.
248  CONTAMINATED AQUIFER CONTROL

-------
       Paper withdrawn by author
While printing of Proceedings in progress.
                            CONTAMINATED AQUIFER CONTROL   249

-------
                      Paper withdrawn by author
               While printing of Proceedings in progress,
250  CONTAMINATED AQUIFER CONTROL

-------
                               Using Risk  Concepts  in  Superfund

                                              Joel S. Hirschhorn, Ph.D
                                                 Kirsten U.  Oldenburg
                                                      David  Dorau
                                          Office of Technology Assessment
                                                United States Congress
                                                    Washington, DC
ABSTRACT
  Using risk concepts to develop and implement policies for Super-
fund is intrinsically attractive. But there are issues and disadvan-
tages that make application difficult and contentious. There is little
to gain unless quantifying risk concepts can concretely improve
decisions. However, it is the attempt to move from simple quali-
tative concepts to numbers that causes disagreement and opposi-
tion to the conclusions the numbers imply. We will discuss several
issues related to using risk-based policies in the Superfund pro-
gram and present their implications for Superfund policies. There
may be more to gain from Qualitative discussions of risks than from
formal risk assessments based on uncertain data and unreasona-
ble assumptions. Thinking about risks may be much more pro-
ductive than calculating them.

INTRODUCTION
  There is increasing interest in the concept of risk in the environ-
mental area. People see risk as a  scientific basis for improved
decision-making. There is a common sense attractiveness to saying
that action should depend on how much risk arises from the action
compared to inaction or other actions. Since reducing risk always
costs money, it seems natural to relate environmental protection
actions to the magnitude of risk present and  to how much risk
reduction is possible.
  It is quite another matter, however, to go beyond these simple
concepts to quantitative risk assessment. One recent analysis said,
"As  might  be expected, the Superfund program is the source of
most of the demand for the use of site-specific risk assessments.
... site-specific risk assessments to  guide the Superfund program
have been  accepted in principle but remain to be fully imple-
mented."1 This may be an understatement of the real world con-
straints  to  using formal, quantitative risk assessments in the
Superfund  program.
  It should be clear from the onset that most ordinary people are
suspicious of risk assessments as a basis for critical decisions. Risk
assessments typically are used to defend a specific action or  deci-
sion, especially in the context of balancing cost against risk. But
the public,  which is primarily interested in being protected from
some specific hazard, is less interested in cost balancing and project
justification than in seeing the hazard removed quickly and
permanently. The public often sees  proponents of risk assessment
as trying to convince people to accept risks that the  proponents
do not face rather than acting to remove them. Moreover, people
are being told to compare unlike risks. But risks vary remarkably
in the degree to which they are voluntary or imposed.  The public,
thus far, has a moral and legal right to accept some risks and reject
others while the government may have a moral and legal respon-
sibility to reduce some risks and not others. Objectively speaking,
this situation  may make no health, environmental or economic
sense.
  Risk assesssment always has been and  will remain intimately
mixed up with ethical and economic issues. Going beyond con-
cepts to formal, quantitative analysis offers the appearance of scien-
tific certainty that exists in the physical sciences and engineering.
But this is generally a facade that quickly crumbles under objec-
tive scrutiny, usually because the data are so poor. Moreover, the
application of risk concepts means making many assumptions and
using many simplifications that implicitly introduce personal or
institutional values into the process.  The  fact is that formal,
quantitative risk assessment is not a science, nor does it offer,
today, a reliable sole or primary basis on which public or private
institutions should make important decisions affecting the general
public. However, risk concepts in qualitative discussions can be
useful to analyze options and make decisions when combined with
other, traditional forms of information. When risk concepts are
openly considered, new issues and solutions may arise simply
because the discussion is focused on effects instead of on historical
factors, responsibility and descriptions of the problem.

SUPERFUND OPPORTUNITIES
  The Superfund program and, indeed,  any other public or private
cleanup of chemically contaminated sites,  offers a number of
opportunities to use risk concepts. First, there is the need to decide
whether a site requires a cleanup. Second, there is the need to decide
exactly what kind of cleanup is technically feasible, economically
acceptable and environmentally sound. This is the remedy selec-
tion stage. Third, there is often the need to decide when a cleanup
is over; that is,  has an acceptable degree  of environmental pro-
tection or risk reduction been achieved?
  The information usually gets better as decisions move along this
spectrum. However, as the information gets better, the decisions
may become less important. An underestimate of risk that results
in a false negative (no cleanup required) at the first stage can be
catastrophic to those  exposed to  the risk. In  the Superfund
program, underestimate of risk could  result from a preliminary
assessment and site investigation that generated poor information
for the Hazard Ranking System (HRS) or  even from a systematic
problem with the methodology of  the HRS. The worst mistake
would be such a low HRS score that the site never gets on the NPL
and, hence, is never considered for a  remedial cleanup.
  At the second stage, the RI/FS can generate a lot of  good
information to assess future risks or risk reductions from several
                                                                                                RISK ASSESSMENT     251

-------
cleanup technologies or plans. The presumptions, however, are that
the information is correct and that the cleanup approach will yield
the results that the study anticipated. Neither presumption is neces-
sarily correct. The very nature of contaminated sites is that they
contain surprises that become evident only upon extensive field
work. But at least a decision will be made to clean up the site, with
the worst case a cleanup that is only partially effective. The chances
are that some, perhaps most possible, risk reduction will occur,
although it might not be as cost effective as anticipated. Or it may
mean that a great deal of risk has been removed, very expensively,
but that  still more  significant risk remains,  which is now pro-
hibitively expensive to remove.
  By the time the third stage is reached, actual field cleanup has
yielded a lot of excellent, site-specific information. Now the criti-
cal decision is  when to stop a particular activity, such as treating
of groundwater, or whether to go beyond the early stages of a
cleanup,  such as excavating soil because not all the contaminants
have been removed  by flushing the site with clean water. A poor
assessment of remaining risk will mean that cleanup will not be
as effective as it could have been. One site has made it to the NPL
a second time because it wasn't really cleaned up the first time.

SCOPE OF  RISK ANALYSIS
  There is always the issue of what risks should be counted. Is
analysis of a specific cleanup to take into account other sources
of risk to the people affected by the cleanup? Possibilities include
other cleanups, hazardous and solid  waste management facilities
(e.g. incinerators), routine and accidental releases from industrial
sites,  and a  host of conventional risks of contemporary life. A
narrow analysis of a cleanup risk by itself may conclude that the
risk is not serious, but it may add to other sources and lead  to a
serious cumulative risk. That is, the new risk may interact addi-
tively or synergistically and result in an unacceptable risk to those
directly affected. Thus far, this appears a major shortcoming in
the Super fund program.
  Generally, the Superfund program takes a very narrow view of
risk; the specific site is analyzed as if the people at risk are exposed
to nothing but the risks from that one site. There may be situa-
tions, for example, where a site's contamination is judged so low
that exposure to a specific chemical or class of chemicals from that
site is deemed acceptable. However,  the same people may be
exposed to a similar risk from an operating industrial plant in the
same area. Is there  a moral or legal need to decide the cleanup
is necessary, or to increase the level of cleanup, in order to avoid
high exposures? This question needs more attention. There are tech-
nical  and economic issues to address when considering whether
actions should be taken at the cleanup site or, perhaps, somewhere
else.

FACING UNCERTAINTIES
  Quantitative risk assessment and analysis has a nagging problem
of reliability and uncertainty. How good are the data? Exactly what
assumptions have been made? How are fundamentally different
kinds of risk  (e.g., acute versus chronic illness versus different kinds
of health effects such as miscarriages) handled? How should the
results be presented  to portray  fairly the inherent uncertainties of
the data, assumptions  and methodology. To what extent  can
cleanup goals and standards be linked to quantified risk results?
Are there standard ways to perform analyses and present results
and, if not,  should  there be?
  The situation has been summed up very well by Evans: "Some
argue that it  is precisely when the uncertainties are large that struc-
tured analysis is most useful. Others believe  that at some point
quantitative analysis becomes wholly uninformative and that it  may
tend to obscure the most important issues and lead to delay in the
implementation of needed controls."2 Whatever one's views are
about the usefulness of risk assessment in the face of uncertain-
ties,  one thing is clear: "Current risk estimates are  fraught with
uncertainty ..current assumptions often are based on little or no

252    RISK ASSESSMENT
data."3
  As soon as risk assessment becomes quantitative, mistrust and
uncertainty enter the picture, with good reason. Although practi-
tioners of risk assessment acknowledge the uncertainties of theft'
work, almost all results of risk assessments are expressed without
explicit and consistent descriptions of uncertainty. All the assump-
tions and uncertainties may be given in the body of a detailed, tech-
nical discussion, but it is the over-simplified result that appear!
in the most readable and influential parts of reports.
  The public does not necessarily trust the "experts." Indeed, risk
assessment experts usually admit that there is a subjective compo-
nent. Hartung has said that "Improvements  in the development
and  application of dose-response relationships can reduce the
subjective component in the risk assessments which are based on
these relationships."4 He did not say "eliminate" the subjective
component. Wilson and  Crouch have said "Some estimates of
uncertainties are subjective, with differences of opinion arising
because there is a disagreement among those assessing the risks."'
  It  should be obvious that the very fact that cleaning up chemi-
cally contaminated  sites is a new technical effort for modern society
means that we are unlikely to have complete and reliable informa-
tion  on which to make risk calculations, Getting more and better
information will take decades, and it will cost enormous amounts
of money.
  Faced with this situation, people will either stop trying to push
risk assessment beyond its current capabilities or they will try to
compensate for the information shortcomings. Those exposed to
risk normally will  favor the former, while those responsible for
making cleanup decisions and providing funds will favor the latter.
Hence, within the Superfund system, we see increasing use of indi-
cator contaminants instead of the full array of site contaminants;
models  for transport and fate of hazardous substances in the
environment instead of actual data; simplifying assumptions and
extrapolations for dose-response relationships; extrapolation from
one type of health effect to others; and little examination of syner-
gistic effects or the effects on particularly sensitive parts of the
population
  Synergistic effects  may  be  particularly  important  because
exposure levels may be low and chronic effects not readily seen
for some time. Exposure to a number of chemicals at low concen-
trations can be quite different than exposures to high concentra-
tions. When concentrations are low, synergistic effects may make
the difference between no health effect and an unexpected one.

INDIVIDUAL  VERSUS AVERAGE RISK
  A  most  difficult public policy choice  is whether to consider
individual risks or to use some sort of average population risk.
This issue inevitably raises questions about balancing amounts of
money spent to reduce risk against its benefits. Is society to limit
its spending on cleanups because relatively few people exposed to
high risk are involved? What are the implications for cleanups in
different geographical, social and economic settings?
  There has always been a bias within the Superfund program,
as evident in the HRS methodology, to consider population risk
instead of individual risk, making it difficult sometimes for rural
sites  to reach the NPL. A few farm people drinking highly con-
taminated groundwater may get less protection than people in an
urban setting exposed to less individual  risk from a contaminated
site.  Or  sometimes the solution is to provide an alternate source
of drinking water rather than clean up a contaminated aquifer that
few people use. However, time has shown how population shifts
and new use of land occurs. Therefore, such a contaminated aquifer
might become much more contaminated and difficult to clean in
the future when the water becomes a needed resource.
  In the case of radon contamination of homes where individual
risks can be enormous, the government has been very slow to act.
In fact, the risks from three unregulated indoor air pollutants
(radon, passive smoking and volatile organics) are each higher than
the total risk of  five regulated  outdoor air pollutants.6  But

-------
hundreds of millions of Superfund dollars already have been spent
on dealing with relatively low risks, but risks that affect medium
to large populations.
  This issue, perhaps more than any other, brings risk assessment
up against economic and political issues. An effective cleanup of
a heavily contaminated site can cost so much that when only a small
number of people are affected the cost per  person—or per life
saved—appears  enormous and unacceptable. And when large
numbers of people are exposed to risk—or to perceived risk—they
will  be more effective than  a handful of people  in getting the
attention of the news media, organized environmental interests and
the government.
CONCLUSION
   The increasing interest in risk assessment is based on the intrin-
sic appeal of relating cleanup actions and spending to environ-
mental risk. But risk  assessment is no panacea for difficult
decision-making  about individual cleanup sites or policy formu-
lation for the entire Superfund system.  No amount of formal, quan-
titative risk assessment can or will make the three basic decisions
easier—deciding  if cleanup is necessary, deciding which remedy
is best and deciding when enough cleanup is reached. But thinking
about risks may make these decisions better. It is possible to manage
risks without calculating them.
   Public  perceptions are key in this area. Slovic  has made impor-
tant observations: ".. .experts appear to see riskiness as synony-
mous with expected annual mortality. ... 'riskiness' means more
 to people than 'expected number of fatalities.' Attempts to charac-
 terize, compare, and regulate risks must be sensitive to this broader
 conception of risk. Each side, expert and public,  has  something
valid to contribute. Each side must respect the insights and intelli-
gence of the other."7 And Lave has said, "Since they are the con-
sumers  and the voters  in our  democracy, people are the final
arbiters of how safe is  safe enough."3
  The day when risk assessment will be as reliable and useful as
a scientific calculation is beyond the horizon. But when it arrives,
there probably still will be cleanups that need risk assessment. And
all the cleanups that will have preceeded that day will have helped
generate the information to make risk assessment the effective tool
that so many people want today. In the meantime, it is not only
possible, it is necessary  to have an effective Superfund program
without accurate and reliable  calculations of risks. Too much
reliance on risk numbers today  may be  counter-productive to
achieving cost-effective  cleanups  before sites get worse.

REFERENCES
1. Russell. M. and Gruber.  M . "Risk Assessment in Environmental Policy-
  Making," Science. 236, Apr. 17, 1987, 286-290.
2. Evans. J. S . Environmental Risk Management. Is Analysis Useful?
  APCA, Pittsburgh, PA 1986, iv-vi.
3. Lave, L B.. "Health and Safety Risk Analyses: Information for Better
  Decisions." Science, 236. Apr., 17, 1987, 291-295.
4. Hartung. R., Environmental Risk Management—Is Analysis Useful?
  Air Pollution  Control Association, Pittsburgh,  1986, 67-72.
5. Wilson, R. and Crouch,  E. A. C., "Risk Assessment and Comparisons:
  An Introduction." Science. 236, Apr. 17, 1987, 267-270.
6. Wallace, L. A ., Environmental Risk Management—Is Analysis Use
  ful? Air Pollution Control Association, Pittsburgh, 1986, 14-24.
7. Slovic. P., "Perception of Risk," Science., 236, Apr.  17,1987, 280-285.
                                                                                                       RISK ASSESSMENT     253

-------
                                        Risk Communication:
               A Critical  Part  of  the  Public  Participation  Process

                                                  Susan L. Santos
                                              E. C. Jordan Company
                                                  Wakefield, MA
                                                   Sally Edwards
                                    U.S.  Environmental Protection Agency
                                           Waste  Management  Division
                                                    Boston, MA
ABSTRACT
  The methodology of risk assessment is used commonly as a
framework for environmental decision-making at the U.S. EPA
and other government agencies. Risk assessment results are used
to establish the need for remedial action and to determine the type
and extent  of remediation.  Throughout the  site investigation
process, citizen and decision-maker concerns are focused on the
potential public health and environmental impacts associated with
a hazardous waste site. Once remedial actions are under consider-
ation, concern again focuses on the site risks and the need to eval-
uate the effectiveness of proposed remedial actions.
  Given these concerns and the emphasis on risk assessment, risk
communication is becoming an increasingly important part of the
community relations process at Superfund sites. In order to ensure
effective and appropriate use of risk assessment in the site remedi-
ation process, increased attention must be paid to risk communi-
cation.  Often  risk assessment results are not  effectively utilized
because of the complex and inherent problems in transmitting this
information to decision-makers and potentially affected citizens.
Approaches for communicating risks are presented,  focusing on
the psychology of communication and the issues involved in trans-
lating highly technical information to a primarily non-technical
audience.

INTRODUCTION
  Environmental risk communication has in recent years become
a topic of great interest among government officials, academicians
and citizens concerned about environmental health impacts in their
communities. Much has been written about risk communication,
generally focusing on techniques to improve the transmission of
risk information. These techniques often fall short of addressing
fundamental failures in the risk communication process as they
focus on the technical aspects of risk information, rather than basic
communication principles. From our work with the  U.S. EPA's
hazardous waste  cleanup program, a highly public program where
risk assessment and communication are an integral part of the
decision-making process, we have had the opportunity to observe
many instances of risk communication failures and successes. These
experiences have convinced us of the essential importance of some
basic communication concepts for successful risk communication.
  In this paper, we will revisit some basic principles of communi-
cations, discuss where environmental risk  communication has failed
to focus and why and clarify what risk information we believe the
public is looking for based on our experiences. The  U.S. EPA's
Superfund program will be the focus of discussion throughout this
paper.  From this analysis, we will provide some suggestions for
more appropriate and useful development of risk communication
strategies.

DEFINITION OF RISK COMMUNICATION AND
FUNDAMENTAL MESSAGES
  Before beginning any discussion of risk communication, it is
important to define the term. In a 1986 report to the U.S. EPA,
"A Review of the  Literature for Communicating  Information
about Health, Safety and Environmental Risks," Covello, et al.,
defined  risk communication as "any purposeful exchange of
information and interaction between interested parties regarding
health, safety or environmental risks"1. Other researchers on the
subject have simplified the definition of risk communication to...
"the way in which people learn about hazards." This second defi-
nition may hint at why there are problems in communicating risks,
as the definition implies risk communication is a one-way process.
Somehow  information is being transmitted from a "source,"
although the source is not explicitly identified in the definition.
This definition also implies that the communication is directed at
a "receptor" .. .the people or public learning about the hazards.
We prefer to define risk communication in Covello et al.'s terms
because  it recognizes that risk communication, in fact any com-
munication, is a two-way process1
  Individuals familiar with communication theory will remember
that the concept of communication is described as a two-way
process involving a source, a channel and a receiver. The source
transmits the message via a channel to a receiver. The receiver must
receive  the information  before communication has occurred.
Defining risk communication in basic communication theory terms
illustrates how communication cannot occur without interaction
between the source and receiver. Covello et al. 's definition of risk
communication is  consistent  with this general definition of
communication1.
  The research by Covello, et al.,1  in reviewing communication
efforts related to health, safety or environmental risk information
indicated  that risk  communication problems  arise from four
primary areas:

1.  Message problems -  e.g. the  limitations  in scientific risk
    assessments
2.  Source problems - e.g. limitations of the individuals who usually
    perform the risk assessments or communicate the results
3.  Channel problems - e.g. limitations in how the risk informa-
    tion  is  transmitted
4.  Receiver problems - e.g. the characteristics or limitations  of
    the intended recipients of the communication
 254    RISK ASSESSMENT

-------
  The results of the 1986 EPA Study1 (substantiated by numerous
other researchers) concluded that, in analyzing these basic risk com-
munication problems, there are four main messages which risk com-
municators should recognize:

• Know your risk communication problem
• Know your risk communication objective
• Use simple and non-technical language
• Listen to your audience and know their concerns

WHAT HAPPENS WHEN  WE COMMUNICATE
RISK INFORMATION
  This paper will take a closer look at the basic recommendations
for risk communication. While we  can  agree with the general
problems cited by Covello, et al.,1 the responses from risk com-
munication experts and researchers regarding solutions to these
problems have serious flaws  in that they emphasize the issue of
risk rather than communication. Researchers and practitioners have
focused too heavily on certain aspects of risk, such as the difficul-
ties and limitations in the scientific process or the perceptions that
risk information invokes within the general public. Although these
unique problems associated with risk communication are impor-
tant, risk communication theory has failed to analyze why these
problems or unique concerns may arise.
  The literature leads us to believe that it is the highly complex
technical and emotional nature of risk issues. We believe that it
goes back to the basics of communication theory, i.e. that the key
to risk communication is the communication process itself.
  How does this relate to the four main messages  for risk com-
municators listed above? From our perspective as  risk assessors
and risk communicators, we  believe too much emphasis has been
placed on two of these messages: (1) knowing your risk communi-
cation problem (2) the use of simple and non-technical language.
These are primarily risk-related issues rather than basic commu-
nication issues, although strictly speaking they are  message or
channel problems. Thus, the emphasis has been on only part of
the communication equation (i.e., the message and  channel). Not
enough attention has focused on the receiver. More specifically,
not enough attention has focused on ensuring that both source and
receiver considerations are addressed in any risk communication
(i.e. that communication is two-way). Relative to Covello, et al.'s
main messages for risk communicators, two of the messages need
more focus: (1) knowing your risk communication objective  and
(2) listening to your audience and knowing its concerns. We need
to better understand why a two-way  process is not occurring.
Individuals communicating risk information often  focus only on
the message and do not consider the conflicting needs of the
receiver.
  We believe that risk communicators must re-examine risk com-
munication  problems in light  of good  basic  communication
practices and needs. Source, channel and receiver all must be con-
sidered to allow a purposeful exchange of information and
successful communication. This improvement in communication
can be accomplished by analyzing the two messages of Covello,
et al. which have not received enough attention: (1) listen to your
audience and (2) know your objective.

LISTENING TO THE AUDIENCE—RECEIVER PROBLEMS
  Most risk communicators  agree that listening to the audience
and knowing its concerns is of fundamental importance. However,
this issue often is addressed superficially and late in the risk com-
munication process. Effectively listening to the audience requires
a serious effort to know the group and its concerns prior to relaying
the risk information. In terms of the Superfund process, this means
interaction must occur long before the public meeting or citizen
briefing where the results of the risk assessment are communicated.
For effective communication to occur, the source must present and
disseminate risk information in a manner which acknowledges and
addresses the receiver's concerns. To do this, risk communicators
must determine: (1) what is relevant to the audience; (2) what in,
formation is necessary to communicate what is relevant; and (3)
how the risk information can be presented effectively to meet both
the communicator's message needs and objectives and the receiver's
concerns.
  Numerous experts on risk communication including Sandman
(2), Fischoff (3), Keeney, et al. (4), Kasperson (5) and Slovic (6)
have described problems that arise in communicating risk because
of the inherent differences between those communicating and those
receiving the risk information. These differences are barriers to
effective listening. To overcome these barriers, a framework for
communication between the source and receptor is necessary to
ensure that we approach risk communication from a basic com-
munication standpoint as well as from a technical risk perspective.
  To achieve effective risk communication,  we must answer the
following questions:

• Is the  communicator listening to the audience and at least
  acknowledging the concerns if he cannot  respond?
• Who is communicating the  risk information and is he able to
  communicate?
• Can the objectives of the risk assessor/communicator be met
  while still meeting the information needs  of the public?

  If the answer to any of these questions is "NO," then regard-
less of how we try to simplify the message,  use comparisons to
other known risks and prepare graphics, etc. to convey risk, we
will not have effective communication (i.e. communication that
meets the communicator's objectives and responds to the receiver's
needs).

Acknowledging What the Community Wants  to Know
  Our experience with the U.S. EPA's Superfund program has
given us some indication of the  kinds of questions community
members ask, issues concern them and answers that satisfy them.
The discussion below summarizes some of the common concerns
we have heard expressed by people in communities affected by
hazardous waste sites.
  A fundamental concern of the public has been termed the micro
risk question, "What does this mean to me,  personally?"7. U.S.
EPA risk communicators have a tendency to focus on the macro
risk picture, i.e. presenting a technical assessment of the potential
threat posed to "public health," a hypothetical audience that is
not specific to the individuals in the community in question. This
macro focus of the U.S. EPA is not surprising, considering the
agency's objectives and the data available from a hazardous waste
site investigation. Generally there are no data available on the in-
dividual health status of community members. Even if there were,
the U.S.  EPA does not consider this information necessary to
determine the scope of cleanup. However, this difference in ob-
jectives and information requirements can create a discrepancy
when U.S. EPA officials and community members attempt to com-
municate. It  therefore becomes the responsibility of the risk
communicator to re-interpret the risk assessment, to find appropri-
ate ways to personalize the site information to provide relevant
answers to the community.
  The community also wants total assurance that its environment
is safe. People are uncomfortable with probablistic statements from
government officials, especially when they feel that they have no
control over the situation2. Again, there  is a conflict between the
information government officials are gathering,  analyzing and
presenting as scientists and  engineers in a complete rational,
accurate and scientific format and what the community wants to
know (i.e'. what  is the bottom line; what is  the absolute).
  A related concern for the public is the question, "What level
is safe?" (can I drink the water, plant a garden, etc.). U.S. EPA
scientists generally evaluate limited exposure and toxicology data
to make this determination. Often, rather than really listening to
the question and give a clear answer based on an informed judge-
ment, the scientist will give a vague, unclear answer which focuses
                                                                                                  RISK ASSESSMENT    255

-------
on the uncertainty of and limitations in the data. To give a meaning-
ful response to be meaningful to the community, the U.S. EPA
must be able to give yes/no answers.
  A final major community concern often is, "I've been drinking
this water for 10 years before you sampled it and told me it was
contaminated. Will  I get cancer?" This historical exposure issue
is a serious one which the risk assessor usually cannot answer in
a satisfactory way (there is usually no way to reconstruct a histor-
ical exposure scenario). Instead, U.S. EPA scientists often pro-
vide answers  regarding present and future risk. This type  of
response falls short of reassuring people regarding their concerns
about past exposure.
   All of these concerns are  a reminder that those affected by the
contamination problem are really the experts in judging the sig-
nificance and implications of risks posed by a hazardous waste site
or other environmental problems8. To effectively communicate
risk, those responsible for communicating must learn to listen to
people in these communities.
Listening to Foster Communication
   Numerous  researchers have provided insight into listening
problems. Slovic6 has provided excellent insight into how people
perceive risk differently. He  has stressed that those who are respon-
sible  for assessing, communicating and  managing risk must
understand the ways people think about and respond to risk. Social
psychology provides insight into the differences between how the
"experts" and "lay persons" view risk. Work by Keeney, et al.,4
describes structuring differences and language difficulties between
the general public and experts. Kasperson5 has related many of
these issues to the  concept of credibility  and trust. Sandman2
refers to experts and policy-makers working at a different level
of analysis than the media or general public. In essence, all these
researchers  are saying that  the communicators  and receivers are
working on different levels. They view risk differently and they
have conflicting objectives/needs; thus a two-way process of com-
munication cannot  occur.
   Listening and communication cannot occur if the source is com-
municating on one wavelength and the receiver on another. The
following examples based on our risk communication experiences
in the Superfund program further illustrate the need for listening
and two-way communication and demonstrate how by listening
and acknowledging the public's concerns we can foster effective
communication.
   At a Superfund site in New Hampshire, citizens were concerned
over groundwater contamination. The key facts in  this example
are that the contamination  was not in the potable water supply,
that no actual exposure had occurred and that only low levels of
volatile organics had been found in the monitoring wells tested.
This information was presented to the public and numerous citizens
raised concerns, asking if the water "was safe to drink" and "could
they bathe in it." The response given was that the level of volatile
organics detected was below U.S. EPA established drinking water
standards set for the protection of public health.
   Further questions from the public focused on the specific levels
of contaminants found and  the fact that several compounds were
detected. They again asked  if it was safe. Again, an answer was
given in terms of U.S. EPA standards and the increased incremental
carcinogenic risk. The public only seemed to get more concerned
and angry over this answer.  The risk  assessor/communicator
became defensive and ultimately a serious problem arose. Clearly,
communication was not occurring.
   The problem was  that the risk communicator did not listen and
did not attempt to relate the technical information he had to
respond to the citizens' needs. By intervening in the situation, we
instead focused on what was being  asked. Instead of responding
in terms of the results of the risk  assessment, we  provided the
following information:

•  Question: Is it safe?
•  Answer: Yes—The water is safe.

256    RISK ASSESSMENT
• Question: But  what about  the chemicals you found—those
  should not be in my water.
• Answer: I understand how you feel. I know you do not like the
  idea of having chemicals in your groundwater. I understand your
  frustration and concern, but if you're asking me if I think the
  water is safe, then in my professional judgement, the answer
  is yes.

  Immediately, the tone of the  audience changed.  The citizeoi
became less defensive because someone listened and answered their
questions instead of just offering the technical information that
we as risk assessors are comfortable with or feel the public needs
to know. Later in the meeting, people were willing to learn about
what U.S. EPA drinking water standards mean and how the ritk
assessment evaluates exposure and risk. However, this interest wai
evident only after their questions and concerns had been answered.
  This example points out the  need to have people who can com-
fortably listen to the public communicate risk information. If scien-
tists and government officials are not comfortable with this, they
should not be the ones communicating.
  The risk communicator must be able to relate to  the public OB
its  terms. Many risk analysts use technical  and   bureaucratic
language when communicating  risk information to the public.
Although this language may be appropriate for the risk assessment
itself and in communicating with other experts, technical language
is not appropriate for risk communication purposes. Most scien-
tists have a tendency to be precise and, as such, they often describe
all the uncertainties and limitations associated  with a risk assess-
ment. This amount of information is frightening  and overwhelming
to the public which is trying to figure out what the risk means to
it; the public wants certainty, not caveats.
  Scientists and government officials need to put their needs aside
to communicate with the public. This admonition is not meant to
be condescending; if anything, it recognizes the legitimacy of the
public's concerns and need for information.
  Risk communicators can communicate better by giving simple,
direct answers, by communicating results on both the  individual
and population levels and by personalizing risk. They need to leave
out the technical qualifiers that professional colleagues insist on
and that academic training has instilled in them. The most appropri-
ate  person to simplify and translate a risk assessment  is the risk
communicator.
  Another example further demonstrates the need  to  listen and
attempt to understand the concerns of the audience.
  At another Superfund site,  residents located near the former
landfill were concerned about children's exposure to contaminated
soils and leachate seeps. The original response was to discuss the
low levels found and provide the  quantitative risk estimates. This
answer did not address the citizen's concerns, especially mothers
who were concerned about risks  to their children. The technical
person explaining the risks tried  to simplify what the risks were
by saying that 3.2 x 10-? was a low risk. This did not seem to
help. It was clear that the site manager was not  comfortable with
trying to allay the fears of the citizens. This is not an uncommon
problem with most risk assessors. The most common communi-
cators of risk  information are professionals who are  trained to
ignore feelings, to not personalize scientific matters and to do their
job without emotion. This personality difference between the scien-
tist  or government official and the public at large may result in
a communication impasse. To deal with this problem may again
mean making  sure that  the right person communicates the risk
information.
  The risk communicator must  be able to acknowledge the public's
feelings and concerns even when  it means giving an accurate but
less than 100% technically precise response. In the above example,
we first acknowledged the mother's concern for  her child's safety.
We then asked her if she thought her child had occasion to play
near the seeps or come in contact  with the area of soil contamina-
tion. From this exchange, it became obvious that exposure was

-------
limited. If her child did not play near contaminated areas, there
would be no risk. Now, this response may sound simplistic, but
it was effective because we first listened and then responded not
in scientific words or with comparisons which attempted to simplify
the risk information, but in the terms of the individual who raised
the question.
  The bottom line is to know your audience, listen to the questions
and then find ways to express the information the audience needs
and wants in clear, concise terms. As scientists, we must not be
afraid to answer questions to which we do not have perfect answers.
must be willing to communicate.

RISK COMMUNICATION OBJECTIVES—
SOURCE/MESSAGE  PROBLEMS
   Covello, et al.'s messages of clarifying communication objec-
tives and listening to the audience are interrelated and must be con-
sidered as a sort of feedback  loop  to result in successful  risk
communication1.
   As Covello, et al. state, risk communication can have a number
of different goals (e.g., information and  education, behavior
change and protective action, disaster warnings and provision of
information or joint problem solving and  conflict resolution).1
The authors emphasize the importance of knowing the risk com-
munication objective. This is problematic in the Superfund program
because of the complexity involved in  meeting the  underlying
objective of the CERCLA legislation. Although the fundamental
objective  of the  Superfund program is to cleanup abandoned
hazardous waste sites,  in actuality this objective is multi-faceted
and comprised of many interrelated  goals.  The U.S. EPA often
has several objectives to meet with the same information and at
the  same forum.  For example, the U.S.  EPA may focus on
providing the  community with  general risk information while at
the same time it is introducing its preferred cleanup alternative9.
   Trying to meet multiple objectives can create difficulties in com-
munication. For example, for each Superfund site, a  risk assess-
ment is developed which describes potential risks to public health
and the environment in  the absence of any cleanup action. In
essence, this document is a planning document used to help develop
cleanup alternatives. Problems arise  because it also becomes the
information to be used in risk communication, absent the trans-
lation and interpretation necessary to  make the document relevant
and of use to the community.
   U.S.EPA policy development and changes also impact and affect
risk communication objectives. For example, SARA calls  for
permanent remedies and a reduction in mobility and toxicity of
contaminants wherever  practicable. This shift in objectives changes
the way risk is evaluated within the agency and thus is communi-
cated in a public forum.
   Since these multiple goals and objectives will not go away at the
U.S. EPA, it becomes the responsibility of the individual risk com-
municator to sort out, clarify and differentiate among these goals
for the purposes of effective risk communication. For example,
the results may be discussed in an internal meeting regarding setting
cleanup objectives. In this situation, the most important goal often
is to explain to agency individuals not educated in risk  assessment
the limitations  of the analysis in setting goals for cleanup. The same
results must then be translated and interpreted to provide relevant
answers to the community where the focus will be on totally differ-
ent concerns. As such, different answers need to  be provided.

Clarifying Objectives to Effectively Communicate
  Rather than attempting to unrealistically simplify the problem,
it becomes necessary to understand the multiple objectives of Super-
fund so that the goal at a particular point in time is clear and allows
for communication to  occur.
  Recently we evaluated a site with high concentrations of PCBs
in soils. Based on the SARA amendments,  we set cleanup goals
for PCBs in soil that would allow residential development of the
area. These target levels were developed using a number of assump-
tions about likely human exposure, such as years exposed, fre-
quency of contact, absorption rate, etc. Once this analysis was
completed, we were faced with communicating the significance of
these numbers to three different audiences, U.S. EPA management,
the PRPs, and the community affected by the  site. Each consti-
tuency had different primary objectives in mind regarding cleanup:
the U.S. EPA wanted to make sure the law was being implemented
properly and consistently; PRPs wanted a cleanup alternative that
they believed was  cost-effective; and the community wanted a
cleanup that would eliminate the risks posed by the site. As risk
communicators, our objective  was to  present the information
clearly, acknowledge its limitations and use our own understanding
of risk to find a means of translating the information that would
be satisfactory to all three constituencies. We dealt with these differ-
ent objectives by communicating the appropriate risk information
in different forums. Often communication fails  because risk com-
municators try to answer all three objectives with the same infor-
mation no  matter with  whom they are  communicating.  The
communicator must really look at  the objective of the moment,
for that particular situation and translate the information to make
it relevant to the audience in question.
  Again, knowing your objectives and knowing your audience are
intertwined. Understanding the  goals of different groups, as well
as knowing our objective in a particular situation, allowed good
communication and resolution  to occur.

CONCLUSION
  To have  successful  environmental risk  communication,  it is
essential that risk communicators pay closer attention to the com-
munication process itself. Two-way communication involves both
the receiver and the source.  Our experience indicates that for risk
communicators, this  means  understanding your objective and
listening to  and knowing the concerns of your audience.
  Risk communication does  not have to be complicated. The
problems encountered to date  are real, but they can be overcome.
We must first recognize that no amount  of research will replace
the need for basic communication skills. We need to take a hard
look at who is communicating instead of just focusing  on what
is being communicated. There is  no substitute for a good commu-
nicator. No  simplifications, no comparisons and no pat answers
will replace the need to listen, acknowledge concerns and respond
openly and honestly.
  How can we as practitioners unravel risk communication? As
many citizens have expressed  when asked about what they want
from those communicating risk... "I just want someone to listen;
I  just want someone to answer my questions; I just want someone
to care."
REFERENCES
1. Covello, V., von Winterfeldt,  D. and Slovic, P., "Risk Communica-
  tion: An Assessment of the Literature on Communicating Information
  about Health, Safety and Environmental Risks," Draft  Preliminary
  Report to the U.S. EPA, Washington, DC, Jan.  1986.
2. Sandman,  P., "Explaining Environmental Risk—Some Notes on En-
  vironmental Risk Communication," U.S. EPA, Washington, DC Nov
   1986.
3. Fischoff, B., "Managing Risk Receptions," Issues in Science and Tech-
  nol.  11,  1985, 83.
4. Keeney,  R. and von  Winterfeldt, D., "Improving  Risk Communi-
  cation," Risk Analysis 6, 1986, 417
5. Kasperson, R., "Six Propositions on Public Participation  and their
  Relevance  for Risk Communication," Risk Analysis 6,  1986,  275
6. Slovic, P., "Perception of Risk," Science 236, 1987, 280
7. Sharlin, H., "EDB: A Case Study in Communicating Risk," Risk Analy-
  sis 6, 1986, 61
8. Otway, H., "Experts, Risk Communication and  Democracy  " Risk
  Analysis 7, 1987,  125
9. Conn, W. and Feimer, N., "Communicating with the Public on Environ-
  mental Risk: Integrating Research and Policy," The Environ. Prof. 7,
  1985, 39
                                                                                                  RISK ASSESSMENT     257

-------
                            Procedure to  Assist  Decision-Makers
                                                In  Selecting  A
                  Remedial Alternative  for  Hazardous  Waste  Sites

                                                Ram Swaroop, Ph.D.
                                                Richard  Carter, P.E.
                                           Woodward-Clyde Consultants
                                               Santa Ana, California
ABSTRACT
  The paper presents a procedure to help decision-makers select
a remedial alternative based on the criterion of minimum expected
cost, where the total expected cost is obtained by multiplying the
dollar cost for the remedial alternative by the likelihood of dif-
ferent scenarios. Therefore, this procedure includes: the cost items
such as capital costs and operations and maintenance costs; and
non-cost items such as public approval and antipathy and public
policy considerations. Cost  items enter the procedure via dollar
amounts and non-cost items via likelihoods.
  The procedure requires a systematic evaluation of all costs that
are involved in the implementation of remedial alternatives and
the evaluation of the likelihoods of various scenarios based on
available information and professional judgments. Expected costs
of the different alternatives are then used to rank the alternatives.
The ranking of the alternatives helps the decision-maker select an
alternative on a rational basis.

INTRODUCTION
  Selecting a remedial alternative for a hazardous waste site is often
difficult because a balance must be achieved between  a variety of
considerations and concerns, such as the need to minimize cost
while satisfying regulatory cleanup criteria. The selection of an
alternative becomes complex  when it is recognized that there always
will be some  uncertainty, and that even  the most  satisfactory
alternative when implemented may or may not fully resolve the
concerns. An added feature of complexity is derived from the fact
that the seriousness of contamination and threat often are estimated
either from limited or no data and information.
  This paper is addressed to the decision-makers who deal with
RRI/FSs of hazardous waste sites. These decision-makers are either
planners, reviewers, or approvers of work plans and RI/FS reports;
quality assurance managers; or regulators who sign the records of
decision. This paper provides a procedure that will assist decision-
makers in selecting a remedial alternative for hazardous waste sites.
The  procedure integrates the limited information, professional
judgments, and cost and non-cost items in a consistent framework
and  a rational manner. The procedure is illustrated  using an
example of a municipal landfill where off-site migration of leachate
threatens the nearby residential areas and a  public park.

PROCEDURE ELEMENTS
  The procedure outlined in this paper is based on the decision
analysis approach  and utilizes the  criterion of "expected cost."
The expected cost  used in the context of this procedure refers to
cost items, including capital  costs and operation and maintenance

 258    RISK ASSESSMENT
costs, and to non-cost items, including protection of public health
and environment and institutional requirements. Cost items are
included in the procedure via dollar amounts and non-cost item
via likelihoods based on available information and professional
judgments. The  application  of the  procedure depends on an
adequate description of the following elements:

• Nature and  extent of contamination problem and available
  remedial alternatives
• Uncertainties and professional judgments
• Remedial action selection criterion.

Nature and Extent of Contamination and
Available Remedial Alternatives
  The evaluation of the nature and extent of the contamination
problem requires an assessment based on site-specific information.
The problem resolution depends on the list of technically feasible
remedial alternatives available. For both the assessment and the
remedial alternative selection, the description of the problem should
include the description  of site, site setting and site use history;
summary of data and investigations related to the contamination
problem; and the list of remedial alternatives available, consistent
with the environmental, public health and institutional concerns.
  A  remedial alternative is either a single action or a series of
actions. The objective of the decision'maker should be to select
an alternative that will eliminate or mitigate the contamination
threat at a minimum cost with the least adverse public health and
environmental  impacts. However, only when the  alternative is
selected and implemented, can one evaluate the degree of mitiga-
tion achieved, the actual cost and the degree to which the adverse
impacts have been minimized. Since the future is always uncertain,
the selection of a remedial alternative should include a descrip-
tion  of the uncertainties. The inclusion of these uncertainties is
an integral part  of the proposed procedure. The uncertainties
involved in a site-specific problem are estimated either on the basis
of available data or by professional judgments when data art
limited or not  available.

Uncertainties and Professional Judgments
  In the process of assessing contamination and selecting a remedi-
ation alternative  for a site, data are often scarce and important
factors are only partially known or understood; all of the fore-
going introduce elements of uncertainty. In these situations, state-
ments made by professionals with some degree of belief are referred
to as professional judgments. The accuracy of these judgments,
which may never be known, is a function of the degree of corres-
pondence between the actual conditions and the professional's

-------
perception of the site conditions. It is for this reason that profes-
sional judgments  should be expressed in a probabilistic sense,
acknowledging that such judgments are not always correct1. In the
proposed procedure, professional judgments usually are included
in the assignment of likelihood to various factors. These assign-
ments of likelihoods and their values then are incorporated under
the various scenarios to compute the expected costs of different
remedial alternatives.

Remedial Alternative Selection Criterion
  The proposed  procedure  is  based on  a  decision analysis
framework2. This framework includes a dominance principle and
a criterion for a rational selection of the remedial alternatives. The
dominance principle assumes that the decision-makers are capable
of expressing an  order of consistent preferences over a set of
remedial actions if the outcomes of the actions are known1. When
this principle is applicable, the set of remedial actions partitions
into two subsets,  one containing the dominated actions and the
other containing non-dominated actions. For each selected action
in the dominated subset, there is at least one action in the non-
dominated subset which is preferable to the selected dominated
action. Therefore, such partitioning implies that further considera-
tion should be focused only on the actions in  the nondominated
subset.
   The decision analyses further indicate that since all of the true
outcomes of remedial actions can only be known in the future,
after implementation,  the rational choice of action before the
implementation can only be made on the basis of a criterion of
expected utility of actions3.  The  above  three elements of the
procedure were applied in selecting a remedial alternative for the
leachate migration problem at a municipal landfill (REM II Site).

LEACHATE MIGRATION PROBLEM AT
A MUNICIPAL  LANDFILL
   On the park  side of  the landfill (Fig. 1), leachate bleeds from
the site toward the park and homes have been frequently observed.
Remedial action should be taken to mitigate the leachate migration
problem.
                           Figure 1
            Off-Site Leachate Migration Areas and
                  Leachate Collection System

   A  leachate collection system  has been  installed  along  the
 perimeter of the site above  the  park. This system  consists  of
 extraction wells spaced at regular intervals (Fig. 1). The leachate
 is extracted from these wells by submersed centrifugal pumps and
 is stored in tanks. The purpose of this system is to prevent off-site
 migration of leachate. The adequacy of this system depends  on
 the well pump efficiency, location and the design. The chemical
 analysis of leachate has identified certain constituents that are
 potentially hazardous  to  public health  and  harmful to  the
 environment.
  A number of technically feasible remedial actions or technologies
have been identified for the above leachate migration problem.
These are as follows:

  No action
  Al - Pneumatic pump installation
  A2 - Horizontal passive system
  A3 - Vacuum augmentation collection system
  A4 - Containment barrier installation

  Limited information  on  the  leachate bleeds and the leachate
control system is available. Additional data can be accumulated
if such a step is important to the selection of a remedial alternative.
The objective is  to  develop  a rational procedure  to  help  the
decision'maker in selecting a remedial alternative. The following
notations are used for the development of such a procedure:

  T0 - No serious threat
  Tj   Serious threat
  Z0   Bleeds not observed
  Z, - Bleeds observed
  D0 - Collection system adequate
  D, - Collection system not adequate
  Aj - Remedial action
  P  - Likelihood
  C  - Cost of alternative
  ID0 - Additional data indicate that leachate collection
        system is adequate
• ID1 - Additional data indicate that leachate collection
        system is not adequate

  A remedial alternative in the present context  is a sequential
process of implementing the identified remedial actions. Therefore,
the rational procedure should integrate the cost of alternatives to
resolve the problem, the cost of ultimate cleanup of the park and
homes if the alternatives fail and the likelihood that the alternatives
will protect public health and the environment. This is accomplished
by calculating the expected cost of each alternative. The expected
cost is the product obtained by multiplying the likelihood of failure
of the alternative to resolve the problem and the dollar amount
that will be spent to implement the alternative. Thus, the expected
cost in the procedure should be interpreted as a measure of risk
associated with the alternative and not as the likely dollar expen-
diture in the future.

PROFESSIONAL JUDGMENTS IN
THE SELECTION PROCEDURE
  Professional judgments are used in three places within the proce-
dure to select a remedial  alternative: the assignment of likelihoods
to various  factors and  site conditions related to  the  leachate
problem, the assignment of costs to remedial actions identified to
mitigate the leachate migration problem and the assignment of
likelihood to each remedial action to resolve the public and institu-
tional concerns.

Likelihood of Various Factors and Site Conditions
  The project team obtained some information on the leachate
collection system at the municipal landfill site.  This information
was used to assess if the  existing system was adequate  (D,,) or not
adequate (D,). Adequacy of the collection system refers to  satis-
factory  operation of the system as it was originally designed and
located. On the basis of the limited information regarding  the
chemical composition of the leachate,  the team assessed whether
the leachate contamination did not pose a serious threat (T0) or
did pose a serious threat  (T,) to the public health and the environ-
ment. The project team also made observations of the leachate
bleeds. Extensive leachate bleeds were either observed (Z ) or not
observed (Z,,).                                       '
  It will be noted that for each of the two leachate contamination
threats (T0 and T,), there are four mutually exclusive conditions
possible. These conditions are  listed and shown in Fig. 2
                                                                                                   RISK ASSESSMENT    259

-------
    CONDITION Dปlป} Svllflfn
              tlndi Hoi ObMtiMJ
    COMOITIOM OiZt) Sywnt MM
                                  CONDITION Oil.) >r~-> ป•• '
                            Figuire 2
        Mutually Exclusive Possible Conditions of Adequacy of
           Collection System and Observed Leachate Bleeds
       - Collection system adequate (D,,), and no bleeds observed
      ,  - Collection system adequate (DO), and bleeds observed
         (Z,)
• D,Z0    Collection  system  not adequate  (D,), and  no bleeds
         observed (Z^
• D,Z,  -  Collection  system  not  adequate  (D,),  and  bleeds
         observed (Z,).

  The likelihoods  of these  four conditions  may be different
depending on whether there is no serious threat (T^ or there is
a serious threat (T,).  Assignment of various  likelihoods by the
project  team is shown in Table  1. Other likelihoods  were then
computed by the Bayes formula and are shown  in Table 22. The
entries in these tables indicate that neither condition is completely
certain (P = 1 .00) or uncertain (P = 0); the likelihood lies somewhere
between 0 and 1 .
                            Table 1
       Estimates of Likelihoods Given the Seriousness of Threit
Conta
Leachate Poee
Col lect Ion
System
Leach
Bleed
Obeer
(2o>
W.qu.t. p(o?I,
Not
Adequate P(D, ,
inatlon Does Not
Serious Threat (T_)
P(T0) - 0.3

Not Bleed*
ed Observed
d,i
!/Tฐ' '(S?55/Tฐ'
„ , P(D I A 1
Con tan ins 1 Ion POSOB a
Serious Threat (Ti )
P(Ti ป - 0.7

Bleedi Not Bleeds
Observed Observed
( *o) ( X | )
P( D_ZO/TJ ) p( o_x J/TJ )
PID i /T ) PID i n I
  P(I0/Ti)


  P(T,/I0)





  P(T,/I0)


  P(Z,/T0)

  PUj/T,)
                                                                                                Table 2
                                                                               Calculation of Likelihoods of Threat Given the
                                                                                      Observation* In Leachate from
                                                                                         Likelihoods of Table I

                                                                               • 0.75 * 0.15 - 0.90, P(T0)  • O-3
                                                                               • 0.02 ป 0.02 - 0.04, PIT,)  - 0.7


                                                                               - P(T0J0)/P(I0)
                                                                               • Pd./TolPdol/IPd^TolPITofPdoApflTill
                                                                               - (0.090 > O.J)/(0.90 K 0.3  ป 0.04  ป  0.7)  - 0.90(0


                                                                               • 1- PCT/l) • J-0.9040  • 0.0940
                                                                               • O.OS ป 0.05 • 0.10, P(T0) • 0.3
                                                                               • 0.06 ซ 0.9t • 0.9ซ, P(T,1 • 0.7
                                                                      PIT,,/!,) - P(T02,)/P(I,)
                                                                               - P
Poaซ 5ซriou(
Thrซซt (T,)
(Cl
0
(000
it Likelihood
0.05
O.(0
        Installation

    A3   Horlionlal     12.000
        paซalvซ iyatซli
1.0


1.0
        col lect Ion
                                                                            CootalnMnt
                                                                            bซrrl.r
                                                                            lnซtซl lซt Ion
                                                                                         15.000
U.OOO


K.OOO
                                                                                                               IS. 000
O.ซ0


0.50
Dominance Principle Between Actions.
  For the leachate migration problem  at the municipal landfill,
each remedial action "i" involves the implementation cost, C,,
and the protection likelihood, P,. Action "i" is said to dominate
action "j", if P( is greater than Pj and simultaneously C( is less
than C,. This dominance principle was  used  to categorize the ac-
tions into two mutually exclusive groups: one group consisting of
all the dominated actions, and the other group consisting of the
remaining actions. The remaining actions need further evaluation
as remedial alternatives.
  The entries in Table 3 indicate that if the  threat is not serious
(T0), then action A0—no action dominates,  because it costs the
least and it provides the same level of protection as the other
260    RISK ASSESSMENT

-------
actions. However, when the threat is serious (,), dominance of
any one action  is not clear.  Action A4—containment barrier
provides the best protection (P = 0.95), but at a very high cost. It
is seen by comparing A,, A2 and A3 that A, dominates A2 and A3
because both A^ and A3 cost more than A, and provide less pro-
tection  than A,. Therefore,  on the basis  of the dominance
principle, it was  indicated that for a serious threat (T,) situation,
actions Aj  and A4 were the only viable options. Along with A,
and A4, Ao also was considered in the analysis for comparison
purposes and for compliance with the regulatory requirements4-
SELECTION OF REMEDIAL ALTERNATIVE

Identification of Remedial Alternatives
  A remedial alternative  is a  combination of viable remedial
actions. The following alternatives were identified for the leachate
migration problem at the  municipal landfill:

  ALT 1. Action A0 only
  ALT 2. Action A, only
  ALT 3. Action A4 only
  ALT 4. Action A, first, if A; fails then A4
  ALT 5. Action A4 first, if A4 fails then Aj

  ALT 4 and ALT 5 are conditional sequences of remedial actions
A and  A4.  Other alternatives  could  have been identified;  for
example, ALT 6 could be identified by stating that actions A and
A4 will be taken simultaneously. Consideration of the 1 different
alternatives for evaluation is a policy decision. In the procedure,
the  decision-maker  may  consider all alternatives  for  which
implementation  costs and protection likelihoods can  be  clearly
defined.
Expected Cost Criterion for Alternatives
  Any  alternative selected to  mitigate the leachate  migration
problem at the  landfill will  either resolve or fail to resolve the
problem. Because of the uncertainties associated with the problem,
there is no mechanism by which resolution or failure of an alter-
native can be predicted with certainty in advance of implementa-
tion. However,  the cost (C) and likelihood (P) of  protection of
public health and environment were used to calculate the expected
cost of each alternative. The minimum expected cost criterion was
then utilized to provide a rational framework for selecting an
alternative.
Expected Cost of Alternatives and Alternative Selection
  The calculation of expected cost is easily  accomplished by
developing a decision tree. A decision tree indicates by branches
the sequence of events in a decision process, and each branch is
labeled by the likelihood and the associated cost of the event. The
joints of the decision tree branches are called nodes. Expected costs
are shown at each node. For the problem, the decision trees and
various expected costs are shown in Figures 3 and 4. For this
problem, it was recognized that  if every alternative fails, it will
be necessary to excavate the soil in the park and haul and dispose
of it off-site in a proper manner. The cost estimated for such a
cleanup was estimated as approximately $500,000.
                           Figure 3
           Decision Tree for Expected Cost of Alternatives
              When Leachate Bleeds are not Observed
                           Figure 4
           Decision Tree for Expected Cost of Alternatives
                When Leachate Bleeds are Observed
  In the problem, the decision process starts at the left end with
observations on leachate bleeds (Z0 and Zj). Then remedial alter-
natives follow (ALT 1, ..., ALT 5). The seriousness of contami-
nation threat (T0 or T,) then reacts to the alternative chosen. The
action either fails or resolves the public and institutional concerns.
Branches of the decision tree in Figures 3 and 4 are labeled by the
name of the event (fails, resolves); the numbers in the interval (0,1)
show the likelihood of the event conditional on all previous events.
These likelihoods are from Tables 2 and 3. The dollar amounts
are at the extreme right end of the decision tree. If the concerns
are resolved, the dollar amount is 0. If the action fails to resolve
the problem, the site will require final cleanup at a cost of $500,000.
Calculation of the expected cost is facilitated by starting from the
extreme right end of a branch and working back to the left end.
  The calculations  of expected cost values  are shown on the
decision tree nodes illustrated in Figure 3. For example, at the node
of Z? and ALT 3, the expected cost is shown as $77,350. This was
obtained in steps by starting from the extreme right branches. ALT
3 breaks into two branches T: and T0; then Tj and T0 break again
into two branches; A4 fails showing a $500,000 cleanup cost and
A4 resolves with $0 cleanup cost. The expected cost at the T.I and
A4 node is $500,000 x 0.05 + $0 x 0.95 = $25,000, as shown in
Figure 3. Similarly, at the T0 and A4 node, the expected cost is
$500,000 x 0 + $0 x 1 = $0. Then the expected cost of ALT 3
is calculated from these two expected costs plus $75,000, the cost
of A4. Thus the expected cost of ALT 3 is $75,000 + $25 000 x
0.0940 +  $0 x 0.9060 = $77,350. All of the expected costs are
calculated by this method and are shown in Table 4.
                                                                                                  RISK ASSESSMENT     261

-------
                            Table 4
              Expected Cost of Remedial Alternatives
            On the Basis of Leachate Bleed Observations


Leachat. Problem Not Obaervซd(zn) otปซrv. $44,650 5 45.71S
ALT 2 I
ALT ]l
ALT 4 I
ALT 5:
Install pnaumatlc
pumps (A,) $34,800
bซrrlซr (A,) $77, ISO
A! and A4 1C ป, (alls S 9,760
A4 and A! If A4 (alls S7S,96ซ
$197,460
$ 98.91J
$ 44,292
$ (4,1160
  The entries in Table 4 show that minimum expected cost per-
tains to ALT 4. This alternative prescribes that the first action,
A,-pneumatic  pump  installation,  be completed.  If  this action
resolves the public and institutional concerns, nothing further needs
to be done. However,  if it  fails, then  A4-containment barrier
AnAinstallation should be implemented. Note that ALT 5 also uses
the combination of A,  and A4, the same as ALT 4,  but  the
sequence of A, and A4 is reversed. In reversing the sequence, the
expected cost  of ALT 5 increases  to $75,968.  Therefore,  the
minimum expected cost criterion indicates not only the actions that
will be selected, but also the sequence of implementing the actions.

Worth of Additional Information
  Frequently, the decision-maker faces a situation in which reliable
data are lacking and only two choices are available: delay immediate
action and collect reliable data or take immediate action without
collecting additional data. The proposed procedure can help the
decision-maker to evaluate these two choices by calculating the
expected cost of additional data collection. In summary,  the pro-
cedure enables the decision'maker to compare the financial worth
of the additional information with the cost of obtaining  such in-
formation.
  The adequacy of the existing leachate collection system (D0D,)
was evaluated by professional judgments on the basis of available
information. It was  then necessary to decide whether or not to
collect additional data on the existing leachate collection system
before implementing a  remedial alternative.
  The cost of collecting additional data was estimated at $5,000.
It was recognized that uncertainties will still exist even with the
additional data. Because of uncertainties, if the actual system were
                            Table 5
    Estimates of Likelihoods Before Additional Data are Collected
                            Mot Swlow* CT0)
   LBซcfiafซ Control Sytfw
 Actually It    To Bซ Indle.tMl Mot I
                                  0ปtar*.4     Mot Obftorvotf   ObMrv*
                                   (Z,l          CZ0I       IZ,I
                                 P(Z,/T0I-O.I   PIZj/1,1-0.4  Hlt/1t>
 Uซiu.t.     M.,u.t.       ni^lD^^ PCl.OjIOj/T,,)   PIZjOnlO/T,! PI J.OjIO,/!,!
 10.)         I ID.)        - 0.600     - 0.040      - 0.016     - O.MI
P100)-0.?96 Mot MปqU.t. '^(^o10/^1 P(21ฐ0IOI/T<
(ID,) • 0.1W *0.010
P( iO,/D0>"0.2
,1 ..WD./T,, ^Z^^O./T,,
-0.004 -O.OIZ

Hot M.au.t.  Moou.t*

ID,)        (ID0I

          PIIDg/0,1-0.1
                       PIZ^IOg/T,,) '11,0, lฐ,/'o>   "Vl "VV '"iV
                       • 0.015     -0.005       '0.002      -0.090
 PCD^-O.JC*   Mot M.ou.t.     PIZ^ID/T,,! P(Z,0|IO,/TO>   "Vl "VV '"l0! '
             <">,>       - O.I5ป     -0.04)       -0.010      -0.ปIO
                                                                    adequate (D0, the data would indicate sysjem adequacy (ID,,) only
                                                                    80% of the time, and  inadequacy (ID,1) 20%  of time.  If the
                                                                    system were actually inadequate (D,), the data would indicate
                                                                    system inadequacy (ID,) only 90% of the time and adequacy (ID,,)
                                                                    10% of the time.
                                                                      Incorporating  these  uncertainties  and  using the  entries  of
                                                                    Table I, various likelihoods were calculated as shown in Table 5.
                                                                    From the entries of Table 5, three sets of likelihoods needed for
                                                                    calculating the worth of additional data were derived. These like-
                                                                    lihoods are shown in Table 6. The calculation of the worth of
                                                                    additional data by the decision tree analysis is shown in Figure 5.
                                                                                               Figure 5
                                                                            Decision Tree Showing the Worth of Additional Data

                                                                       It will be noted that before additional data are collected, there
                                                                    are only two sets of information: (1 ) adequacy of the leachate col-
                                                                    lection system, D0 and D,, and leachate observations, Z0 and Z,.
                                                                    In the calculations performed in the  absence of additional data,
                                                                    the likelihoods of T0 and T, given Z0 and Z, were used to calcu-
                                                                    late the expected cost. With the benefit of additional information,
                                                                    the case likelihoods of T0 and T, given D^, D^,, D,Z0 and D,Z
were derived, and minimum expected costs were calculated for each
condition by the procedure described.  These expected costs for
associated observations are shown at the extreme right end of the
decision tree of Figure 5, The associated likelihoods on the branches
of the decision tree are from Table 6. Proceeding from the extreme
right end of the decision tree and moving toward the left end, the
expected costs at the branch nodes are calculated. These costs are
shown on the decision  tree of Figure 5.
   Referring to the two  nodes at the extreme left of Figure 5, the
one with no data collection displays an  expected cost of $34,001,
and the one with data collection shows an expected cost of $33,999.
Therefore, the worth of the additional information, at an expense
of $5,000 for data collection is  the difference  between the two
expected costs, which is $2.00. This is a loss of $4,998. Therefore,
a rational decision was made not to collect any additional data,
and to select a decision on  the basis of available information.
DISCUSSION AND CONCLUSION
  There are several different items that the decision-maker must
evaluate  when selecting a remedial alternative:

•  Will the alternative meet  the  institutional  and  regulatory
  standards and  will it be acceptable to the public;
 262    RISK ASSESSMENT

-------
                             Table 6
               Calculation of Three Sets of Likelihoods
     ID0 - Data Indicates Leachate Collection Systemm is Adequate
    ID, - Data Indicates Leachate Collection System is Not Adequate
      Set 1:   Likelihood of ID0 and IDj


             PdD0)  .  (P(ID0,D0)
                   -  (PdD0/D0)P(D0)
                   •  0.8 ป 0.296  + 0.1 x 0.704 - 0.3072
             PdDj)  -  1   0.3072 - 0.6928


      Set 2;   Conditional Likelihood of Dornj Given IDn and IDj


             P(D0/ID0) - P(ID0/D0)P(D0)/P(ID0)
                     - u.B x u.296/0.3072  • 0.7708
             PIDi/ID,,) ซ 1  0.7708 - 0.2292
             PIDj/IDj) - P(ID1/D0)P(D0)/P(ID1)
                     • 0.2 x .296/0.6928 - 0.0855
             P(D0/IDj) • 1  0.0855 - 0.9145


      Set 3;   Conditional Likelihoods of Leachate Observations  (Zn,Zi)
             Given Combinations DQ!DQ, DjIDg, UQ!D^ and D^ID^
             P(Z0/D0,ID0) = P(Z0D0ID0)/P(D0,ID0)

                         [P(Z0D0ID0/T0)P(T0) +
                        -  (0.6 x 0.3 + .016 x 0.7)/
                          0.2368 - 0.8074
                        -  i   0.8074 ป .1926
      Similar Calculations Show:
            P(Z0/Dj,ID0) - 0.0838, P(Z1/D1,ID0)  - 0.9162
            PIZp/Dn.IDj) - 0.8074, f(Z1/D0,ID0)  - 0.1926
            PIZQ/Dj.IDj) - 0.0838, PIZj/Dj.IDj)  • 0.9162
• What is the financial risk, and should the alternative be imple-
  mented now or delayed until additional information is obtained?

  The discussion that follows focuses on how to use the procedure
to evaluate some of these items.
  For the site-specific conditions, it is not possible to predict in
advance  if the selected alternative (design, operation, sequence)
will meet relevant federal, state and local environmental and public
health standards4. However, professional judgments can be made
to estimate the likelihood that a given alternative will meet a certain
standard.  Similarly, professional judgments  can  be made  to
estimate the likelihood of acceptance of rejection of the alterna-
tive by the public and environmental advocates. By adding a branch
for each item to the decision tree, these likelihoods can be included
into the calculation of expected cost and therefore in the selection
of an alternative by the minimum cost criterion. In these situa-
tions,  it is stressed that the estimates of various likelihoods by
professional judgments be based on the reality and not on the
expediency of the situation.
   The financial analysis based  on present worth costs, return on
investments or such methods should be entered into the expected
cost calculations by the actual dollar amounts. The decision-maker
can  use these dollar amounts in the calculation of expected cost
to evaluate either the sensitivity or the worth of various items in
the selection of an alternative.
   The usefulness of the proposed procedure depends on the appro-
priate integration of all the elements that influence the selection
of a remedial alternative.  This  is not easy to accomplish because
data usually are limited and professional judgments can be biased.
The proposed procedure used the minimum expected cost criterion.
This criterion is a reasonable one. However, in a specific problem,
other criteria such as minimax  or weighted expected loss may be
more suitable. These items should be discussed and resolved at the
policy level before using the procedure. In conclusion, a rational
procedure may be the only way to select remedial alternatives where
multiple objectives need to be satisfied and competing vested groups
or interests are involved.


REFERENCES

1. Hogarth,  R. M., Judgement and Choice.  New York, John Wiley,
   1980.
2. Ang, A. H-S, and Tang, W. S., Probability Concepts in Engineering
   Planning and Design. New York, John Wiley, 1975.
3. Keeny, R. L. and Raiffa H., "Decisions with Multiple Objectives: Pre-
   ferences and Value Tradeoffs.
4. Guidance on Feasibility Studies Under CERCLA. Washington, D.C.,
   U.S.  EPA, April 1985.
                                                                                                         RISK ASSESSMENT     263

-------
     International  Study  of the Social,  Psychological  and  Economic
                       Aspects  of  Problem  Hazardous  Waste  Sites
                                                   Michael A. Smith
                                                     Sara B.  Baker
                                             Bostock Hill & Rigby Ltd.
                                                      Birmingham
                                                    United  Kingdom
ABSTRACT
  Studies into the social, psychological and economic aspects of
waste treatment and disposal facilities and of contaminated land
were carried out in 1986 for the European Foundation for Improve-
ment of Living and Working Conditions (an agency of the Com-
mission of the European Communities). Reference to case studies
in Europe showed that there can be significant social, psychological,
and economic impacts on individuals and communities. Reference
to European and North American practice was made to illustrate
how these effects can be mitigated by adoption of public relations
and involvement programs and also how such programs can help
to gain acceptance for operating and planned waste management
facilities. More extensive and detailed studies will be started in the
Autumn of 1987. It is hoped thai these will lead to the prepara-
tion of guidance for local government officials and other involved
with problem sites or siting proposals.

INTRODUCTION
  To find oneself living on or close to a problem hazardous waste
site or other contaminated site can be a traumatic experience. Even
if there are no demonstrable adverse effects on health, there can
be severe psychological pressures and marked social and economic
impacts on individuals and communities. Although such impacts
cannot be eliminated, they can  be  mitigated if those responsible
for the technical aspects (site  identification, assessment and remedi-
ation) are aware of the wider impacts and if the questions of public
involvement, communications and  health appraisal are treated as
integral parts of the problem solving process. It was hard lessons
learned at sites such as Love  Canal and Lekkerkerk that led to the
development of the Superfund Community Relations program in
the United States' and to systematic programs for public involve-
ment and for health surveys in the Netherlands2.
  It was in recognition of the importance of these social impacts.
and of the  important role  of public  involvement in  managing
incidents, that  the European Foundation for  Improvement of
Living and Working  Conditions  commissioned a preliminary
study3  of the social, psychological  and economic aspects of con-
taminated land in  1985 (a parallel study on the siting of waste treat-
ment facilities was also commissioned14.
  This initial study, based on reviews of the literature, discussions
with experts and the preparation of a number of case studies, was
intended to serve as the basis for the more detailed, nation by nation
study,  to be undertaken in  1987-89.
  It is worth pointing out three important features of the separate
contaminated land and waste treatment facilities reports:
• The  first report adopted the definition of contaminated land pre-

264    RISK ASSESSMENT
  viously used in the NATO Committee on Challenges of Modem
  Society (CCMS)  Pilot Study on  Contaminated Land', i.e.,"
  land that contains substances that, when present in sufficient
  quantities or concentrations, are likely to cause harm, to man,
  to the environment, or on occasions to other targets"  This
  definition embraces not only uncontrolled hazardous waste sites,
  but also land that has become contaminated due to past indus-
  trial use, agricultural use, aerial deposition, etc. The definition
  also implies that contamination as such is not a problem: it can
  only be defined as a problem following a full site appraisal taking
  into account site specific factors including current and planned
  future use of the site.
• Although the emphasis in  the second report was on facilities for
  disposal and treatment of hazardous wastes, this arbitrary demar-
  cation, largely  invented for administrative convenience, was
  avoided as far as possible. Hazard depends on the circumstances
  and all wastes are hazardous under certain conditions when the
  full range of possible physical, chemical, biological and radia-
  tion hazards are taken into account.
• For the purpose of the contaminated land report, two types of
  site were distinguished:

    Problem or  potentially problem sites, i.e.,  those  already
    causing or threatening damage and which may, or may nol,
    have already been developed for residential or other purposes
    (the  classic Superfund  site)
  - Those sites where development has been, or is about to be,
    deliberately carried out

  Those engaged in the management of uncontrolled hazardous
waste sites in the United States will be all too familiar with sites
of the first type but generally will,  be less familiar with those of
the second type, a particular feature of the UK scene.
THE  El ROPEAN FOUNDATION
  The European Foundation for Improvement of the Living and
Working  Conditions is a body funded by the Commission of the
European Communities. It  is ruled by a Council comprised of
representatives of member governments, industry and trade unions.
As its name implies, it is concerned with research related to the
welfare of workers and the general public within the member coun-
tries. Thus, while other parts of the CEC administrative machine
concern themselves with policy and technical issues relating to con-
taminated land and hazardous wastes, the Foundation is concerning
itself  with worker safety in the wastes industry and the social,
psychological and economic  aspects  etc of the waste  cycle in
fulfillment of a commitment to the European Parliament to pay

-------
increased attention to environmental issues.

THE PRELIMINARY STUDY ON CONTAMINATED LAND
   Space does not permit a full summary of the preliminary study
(a brief account has been published elsewhere6 and the contents
are summarized in Tables 1 and 2) but a few points are worth noting
because they highlight the differences between the USA and the
European countries covered by the study.
                            Table 1
       Summary Contents of Report on Social, Psychological
           And Economic Aspects of Contaminated Land

 1.   Contaminated land as a technical problem.

 2.   Identification of Contaminated land.

 3.   Case Studies  - (i)  new  studies
                    (ii) from the literature

 4.   Perception of risk.

 5.   Risk analysis and assessment.

 6.   Social and psychological  impacts.

           (i)     social impacts
           (11)    psychological  impacts
           (ill)   effects on  families
           (iv)    effects on  communities
           (v)     disagreements  between citizens  and  authorities
           (vi)    indirect impact on communities

 7.   Medical aspects

           (i)     population  health surveys
           (ii)    role of health professionals
           (iii)   occupational health

 8.   Informing and Involving  the  Public.

           (i)     benefits and objectives of public participation
           (ii)    US policy  and  experience
           (iii)   Netherlands policy and experience

 9.   Economic Aspects.

           (i)     'Problem1  sites
           (ii)    deliberately developed sites
           (iii)   costs to individuals of problem sites
           (iv)    costs of reclamation and development
           (v)     minimising  costs
           (vi)    indirect costs
           (vii)   cost at a national level.

 10.  Centres of Expertise

 11.  Case Studies
   First, not all countries have discovery programs such as the U.S.
Superfund program. Only the Netherlands and Denmark have com-
pleted nationwide surveys7. Partial surveys have been carried out
in the Federal Republic of Germany and only limited surveys in
the United Kingdom. In the UK, site generally are dealt with only
because someone wants to put them to a new, economically bene-
ficial use and, consequently, problems arising from contamination
are dealt  with alongside  engineering problems and within com-
mercial or public finance constraints. In contrast, in the case of
a discovery program such as Superfund sites are tackled as environ-
mental problems.  How a piece of land is recognized as con-
taminated will influence the impact on the community. The UK
type of approach leaves less scope (or perhaps need?) for impact
management.
  Second, attention already has been drawn to the  wide defini-
tion of contaminated land that was employed.  This definition is
essential because much of the land available for development for
residential purposes in the UK and other European countries is
recycled industrial  land. This need to recycle land, and the near
                              Table 2
         Summary Contents of Report on Social, Psychological
    And Economic Aspects of Waste Disposal and Treatment Facilities

 1.   Waste Management Policies

 2.   Haste Disposal and Treatment Facilities

 3.   Environmental Impacts

 4.   Risk Analysis

 5.   Perception of Risk

 6.   Not In My Back Yard ("NIMBI")

          (i)    major Issues relating to  siting proposals
          (Ii)   public confidence

 7.  Social and Psychological Impacts

          (i)    social impacts
          (11)   psychological aspects
          (ill)  disagreements between citizens and authorities
          (Iv)   inappropriate location of facilities

 8.  Medical Aspects

          (1)    the health and safety aspects of wastes
          (ii)   population health surveys
          (ill)  methods of assessing  health risks in  populations living
                near hazardous waste sites
          (iv)   role of health professionals

 9.  Economic Aspects

          (i)    siting hazardous waste facilities
          (11)   Ontario Waste Management Corporation
          (111)  costs to individuals
          (iv)   waste disposal costs in the  United Kingdom

 10.  Overcoming "NIMBY"

          (1)    communication
          (11)   education
          (111)  waste disposal  and land reclamation
          (iv)   recycling
          (v)    compensation

11.  Public Consultation and Involvement

         (1)    benefits and objectives of public  participation
         (ii)   US  EPA guidance  - RCRA Permitting  Programme
12. Case Studies
 impossibility of long-term sterilization in densely populated areas,
 does not appear to have yet had a widespread impact in North
 America, although it can impose severe planning and cost con-
 straints on the revitalization of older urban areas.
    In preparing the report, reliance had to be placed mainly on a
 review of the published literature. Among the important sources
 used in the preparation of the initial reports were accounts of The
 U.S. EPA Superfund Community  Relations  Program1 and the
 Netherlands public involvement programs2. Reference also was
 made to research in the Netherlands on social impacts8  and to
 systematic  programs  for  health  appraisal  developed  in the
 Netherlands9. Resources did not permit first-hand assessment of
 how the various public involvement programmes that were reviewed
 worked in practice. However, it was possible to illustrate the various
 social, psychological and economic impacts  by reference to a
 number of new case studies prepared in the course of the study.
 Three of these case studies are described briefly below. Of these,
 Thamesmead, the technical aspects of which were described in a
 paper presented  at  this  conference in 1984, is probably of the
 greatest interest, as it  is the prime example in the UK of a con-
 taminated site deliberately  developed for housing purposes".
    The cases of Loscoe and Wivenhoe are very small compared to
 Love Canal, Times Beach or Lekkerkerk,  but they nevertheless
 illustrate well the various pressures on individuals and communi-
 ties affected by such incidents.
                                                                                                       RISK ASSESSMENT     265

-------
CASE STUDIES

Deliberately Developed Sites
  In  contrast to the marked and warranted  reaction in the
Wivenhoe and Loscoe cases described below, other major contami-
nated sites have been successfully developed with little or no public
concern.  Such  developments  as  Beaumont   Leys10  and
Thamesmead3'"  were developed in such a way as to eliminate any
risks to the health and safety of the local residents. Given that the
authorities have  exercised proper care, there is no reason for the
subsequent residents to be concerned. Indeed, in many cases, they
will not be made aware of the fact that they are living on a (reme-
diated) contaminated site. However, as illustrated by the case of
Thamesmead described below,  there can  be  pitfalls  in  this
approach.

Thamesmead
  Thamesmead occupies 1600 acres on the south bank of the River
Thames to the east  of London. A large part of the  land was
occupied by  the Royal Arsenal  until  1975.  The contamination
resulted from more than 200 years of use for the manufacture of
armaments.
  The low-lying marshy ground was progressively filled with several
feet of mixed industrial wastes resulting in extensive contamina-
tion with toxic metals, asbestos, the by-products of the production
of coal gas and  electricity,  combustible materials and  methane-
generating materials. Contamination  was first  recognized as a
problem in 1975, prompting a progressive program of site investi-
gation and remediation consisting of covering the contaminated
areas with approximately 3 ft. of clean fill.
  This exercise was generally successful, but changes during 1986
in ownership and administrative arrangements have prompted
requestioning of the  technical basis of the reclamation  program.
In addition,  the problem was not discovered until after  some
development had been completed. At the time  (1975/6), it was
decided as a matter of policy not  to investigate these  already
developed areas, but pressures from residents resulted in a pre-
liminary investigation on  these  areas in  1986.  The site is also
important because, in order to dispose of surplus contaminated
soil, two licensed hazardous waste  disposal sites were created behind
a newly constructed river wall (part of new Thames flood preven-
tion measures). This land has been reclaimed and there are now
houses on at least part of it.
  The Arsenal was operated by  the Ministry of Defence which,
on closure, decontaminated it in terms of clearing all explosive
materials. It then passed to the Greater London Council (GLC)
for comprehensive development  for residential,  commercial and
industrial use. The contamination was only discovered after the
GLC assumed ownership. The original intention was that most of
the housing would be owned by the GLC, but a change in govern-
ment policy  has resulted  in  large areas  being sold  to  private
developers. With the  abolition of the GLC in 1986, control passed
to the Thamesmead  Trust, a  body controlled by locally elected
representatives outside of the normal local government arrange-
ments. Most of the contaminated  area lies in the London Borough
of Greenwich which, since the demise of the GLC, has sought
greater control over  the development  in terms of public health.
It now insists that residents be made aware of the nature of the
land and that long-term monitoring be established to ensure that
the remedial measures are working. There is no evidence that this
is in any way inhibiting sales  of land to developers or, in due course,
of houses.

Loscoe, Derbyshire
  In March 1986, a bungalow was destroyed and three residents
injured in an explosion caused by landfill gas from a nearby refuse
site. A peacetime emergency was declared by the County Council,
and residents of affected homes were evacuated. Emergency work
included the digging of trenches 10 to 13 ft. deep and lining  them
with plastic around the tip, the installation of gas extraction wells

266    RISK ASSESSMENT
in the tip, the establishment of monitoring standpipes in gardeni
and the knocking out of ventilation bricks in the buildings. Mobile
homes were provided for the evacuated residents close to theb1
houses and belongings. Three hundred fifty homes were put on
standby alert, a mobile police station was set up and police patrols
were mounted. No smoking signs were displayed across the road.
Property values suffered.
  The possibility of another explosion was the immediate and main
concern. Feelings  of frustration and anger soon followed when it
become apparent  that the incident  would not receive emergency
action equivalent  to an earthquake response, which residents felt
was justified. An  Action Committee was set up within 2 weeks of
the explosion. Demands were made for a public enquiry, for the
complete removal of the dp, for an independent survey of the health
of the local people and for compensation from the responsible
party. A non-statutory County Council Public Enquiry opened in
November 1986, 8 months  after the problem first arose.
  An interim report issued in June 1987 suggested that the removal
of the tip was not a feasible option and that more gas extraction
weUs should be installed to increase the ventilation of the site. The
local District Council was to purchase the two bungalows most
severely affected  by the gas migration, including that belonging
to an evacuated family who had spent 15  months in a  caravan.
The post-blast price was estimated  at approximately half that of
the original market price. The purchase of other properties adjacent
to the tip is still uncertain.
Wivenhoe,  Essex
  The events surrounding a housing estate in Wivenhoe, built on
unstable and contaminated ground, illustrate how public responses
can be influenced by financial factors. Greater importance seemed
to be attached to the provision of compensation than to health and
safety. An investigation was initiated when cracks appeared in die
walls of houses and the drains failed. Test boreholes by a consulting
engineering firm appointed by the residents' insurance company
revealed that 26 properties, built on the site of a pig farm, were
cracking  up as the ground  below the foundations moved and
settled. A serious health hazard from the farm effluent at founda-
tion level also  was detected.  Despite these findings, the majority
of residents chose to stay in  their homes and to combine to seek
compensation. They feared being left with property that they could
not sell. They also wanted to protect their belongings, fearing theft
and vandalism if they left, and to avoid disruptions to everyday
life such  as schooling.
  This site is also of interest because of the confusion caused by
differences in opinions expressed by various experts. Following the
investigation by Bun early in 1978, the National House Builders'
Council (NHBC) surveyed the remaining homes on the estate (the
NHBC is a mutually funded guarantee scheme provided to pur-
chasers of  new houses by builders to provide insurance against
structural failure). The separate consultants  acting for the NHBC
and for the residents failed to reach full agreement about the likely
behavior of the soil at the bottom of the site but did agree that
there was no risk of a sudden landslide. Reassurances that there
was no risk to public health also  were given. The consultants'
report, however, caused confusion  and bitterness among the
remaining residents (i.e., those living adjacent to the problem area;
those directly  affected had moved out). They chose to continue
to fight for compensation, feeling themselves to be in a worse finan-
cial situation than those whose houses had  been declared unsafe
and who had already recovered their  losses  from insurance com-
panies and the NHBC. They had to decide whether to stay in a
blighted area or to sell at well below market value.
  To alleviate this distress and uncertainty,  the NHBC reclaimed
the estate by buying up the unsafe and evacuated homes and under-
pinning and renovating them. Fifteen houses were treated during
1981, and within a year most had  been sold at about  2000 below
the market value.

-------
THE NEW STUDY

Structure
  The preliminary reports were prepared by the authors alone. In
the next stage, separate reports embracing siting proposals for new
facilities, operating facilities, closed facilities and other forms of
contaminated sites are to be prepared  for  seven  countries by
separate contractors (at the time of writing—July 1987—these have
yet to be appointed). Once these are completed, (the target date
is June 1988), a consolidated European Community report will be
prepared by the  present  authors who also will  be involved in
co-ordination of  the overall project.

Scope
  The separate national reports will cover much of the same ground
as the preliminary reports (Tables 1 and 2) but will concentrate
on what is happening in practice and what research is in progress:
the definition of the various subjects and the overall background
will  be provided by  the preliminary reports. One important
additional area that it is intended to cover is insurance and other
steps that prudent individuals can take to protect themselves against
the economic impacts they might encounter  (e.g., loss of house
which may at present be insured against structural damage but not
against becoming unhabitable because of toxic substances in the
garden).

Ultimate Objectives
   There are a number of important long-term objectives:

• To create a greater awareness of the wider issues among  aH those
   involved in dealing with problem sites, the operation of existing
   facilities and for the siting  of new facilities
• To demonstrate the impacts through selective case studies to show
   that real  people suffer  real effects
• To demonstrate the necessity to increase public awareness of the
   waste issue in general by appropriate education and involvement
   programs; also,  to show the importance of communication
   processes as a way to identify the needs of the public involved
   in such an issue
• To identify research needs
• To provide guidance to those responsible for dealing with such
   matters (usually in local government) on how to manage contami-
   nated land incidents and problems concerning  operating or
   planned  waste  disposal and treatment facilities

CONCLUSIONS
   Experience in North America and in Europe shows that there
can be serious social, psychological and economic impacts from
problem contaminated sites but that these can be mitigated to a
certain extent by properly planned public involvement programs.
Similarly, the problems of siting new waste disposal and treatment
facilities, and the operation of existing facilities, can be reduced
by addressing legitimate public concerns.
  It is hoped that the new European Foundation study described
here, by drawing attention to what has happened in practice (both
good and bad examples) and by bringing together descriptions of
good practice, will lead to a wider recognition of the non-technical
aspects  and lead in due course to authoritative guidance for those
who have to deal directly with such problems.


REFERENCES
 1.  "Community Relations in Superfund: A  Handbook,"  Office of
    Emergency and Remedial Response, U.S. EPA, Washington, DC, 1986.
 2.  Eikelboom, R.T. and von Meijenfeldt, H., "The Soil Clean Up Oper-
    ation in the Netherlands: Further Development after Five Years of
    Experience," Assink, J.W. and van den Brink, W.J. Eds, Contami-
    nated Soil, Martinus  Nijhoff, Dordrecht.  The Netherlands,  1986,
    255-267.
 3.  Smith, M.A. and Baker, S.B., Social, Psychological and Economic
    Aspects of Contaminated Land, Bostock Hill & Rigby, Birmingham,
    UK, 1986
 4.  Smith, M.A. and Baker, S.B., Social, Psychological and Economic
    Aspects of Waste Disposal and Treatment Facilities, Bostock Hill &
    Rigby, Birmingham, UK, 1986.
 5.  Smith, M.A., Ed., Contaminated Land: Reclamation and Treatment,
    (Pterntm, London, UK, 1985).
 6.  Smith, M.A., Baker, S.B. and Pope, W., "The Public Perception of
    Risks Associated with the Reuse of Waste Disposal Sites" Proc.
    Safewaste, Cambridge 1987 (Industrial Seminars, Tunbridge Wells
    1987) 219-224.
 7.  Harris, M.R., "Recognition of the Problem," in Reclaiming Contami-
    nated Land, Cairney,  T.C., (Ed., Blackie, Glasgow, 1986. 1-29.
 8.  De Boer, J., "Community Response to Soil Contamination" in Con-
    taminated Soil, Assink, J.W. and van den Brink, W.J. (Eds., Martinus
    Nijhoff, Dordrecht, The Netherlands,  1986) 211-219.
 9.  Sangster, B. and Cohen, H., "Medical Aspects of Environmental Pol-
    lution: Environmental Incidents in the Netherlands", 1980-1984,
    Clinical Toxicology 23, 1985, 365-380.
10.  Keeps, K.D., "The Reclamation of a Disused Sewage Works" Public
    Health Engineer,  10, 1982, 213-4 & 218.
11.  G.W. Lowe, "Investigation of Land at Thamesmead and Assessment
    of Remedial Measures to Bring Contaminated Sites into Beneficial
    Use." Proc. Fifth Conference Management Uncontrolled Hazardous
    Waste Sites, Washington 1984, 560-564.
                                                                                                      RISK ASSESSMENT     267

-------
                Management  of Waste  Compressed  Gas  Cylinders
                                                    Charles Mattern
                                                      Dan Nickens
                                            Earth Resources Corporation
                                                    Orlando,  Florida
ABSTRACT
  Professionals in the field of hazardous waste management are
increasingly recognizing the difficulties associated with the handling
of waste compressed gases. The problems stem from the chemical
and physical hazards associated with pressurized materials. The
common uses of gases distributed in pressurized containers increase
the likelihood of encountering these types of wastes.
  This paper briefly addresses some of the complications which
can be anticipated when dealing with compressed gas cylinders.
Many of the obstacles relate to inability to  safely transport or
dispose of the gases at permitted facilities. A special category of
compressed gas cylinders, those in deteriorated condition (or whose
valves are inoperable) and whose  contents are unknown,  is also
discussed.
  A case history describing a technology to assist in managing these
cylinders  is also  presented.  A device for sampling and  recon-
tainerizing these cylinders was used on a project for the Corps of
Engineers at a waste cylinder burial site. This project and proce-
dures employed in handling these wastes  are offered as an example
for similar situations.

INTRODUCTION
  Gases,  in compressed or liquified form, are prevalent in indus-
trial and govern, mental applications. Compressed and liquified
gases are also commonly used by  most  households. An example
is the use of propane cylinders  for outdoor gas grills and home
heating. Gases are required in hospitals for medical purposes (i.e.,
oxygen and anesthetic gases) and agriculture employs gases on a
routine basis (i.e.,  anhydrous ammonia for crop fertilization).
Almost every commercial building  is required to have fire ex-
tinguishers containing compressed gases for chemical and electri-
cal fires.  Most refrigeration systems use compressed gases.
  A compressed gas is defined by the American Society for Testing
and Materials  (ASTM) as any material  or  mixture having an
absolute pressure exceeding 40 lb/in.2 at 70'1- Regardless of the
pressure at this temperature? a compound which has a pressure
exceeding 104 lb/in.2 at 130ฐF or  any liquid  flammable material
having a vapor pressure exceeding  40 lb/in.2 at 100ฐF is classified
as a compressed  gas.
  There are two major categories of compressed materials: (1) com-
pressed gases  and (2) liquified  compressed gases.   Liquified
compressed gases condense at normal ambient temperatures under
pressure of 2000 lb/in.2 or less.
  In addition to these major classes, gases generally can be divided
into loosely knit families related by common origins, properties
or uses. These families include atmospheric gases, gases produced

268    VOLATILE ORGANICS MONITORING REMEDIATION
through fractionation (argon and the rare gases), fuel gases,
refrigerant gases and poison gases.
  The hazards associated with compressed gases come in many
forms.  Almost any gas can become an asphyxiant by simply dis-
placing the oxygen in air.  Other gases have toxic effects either
through inhalation or contact with skin or eyes. Obvious hazards
associated with flammable gases include fire or explosion. Special
hazards are associated with gases that are reactive.
  The  very nature of compressing  these compounds  generates
physical hazards  from  possible rupture  or other  uncontrolled
release. A 1.5 ft3 cylinder pressurized to 2.000 lb/in.: contains an
energy  equivalent to 1.4 Ib of TNT.  If a cylinder valve is broken
a very  dangerous projectile can be created.
  In recent years managers dealing  with hazardous wastes  have
encountered an especially difficult problem associated with gas
cylinders. In some instances, the containers may be in deteriorat-
ed condition. The condition of the cylinders may make them unsafe
for transportation. The contents of these dilapidated cylinders are
often difficult to identify. Because the materials cannot be safely
handled, transported or disposed of without'proper identification,
the  options for managers of these wastes are extremely limited.
  The specialized problems associated with compressed gas con-
tainers in deteriorated condition or whose valves may be inopera-
ble has generated a unique technology to overcome the potential
difficulties. This technology, involving a cylinder sampling and
recontainerization system,  was used at a waste cylinder disposal
site in  Jeffersonville, Indiana. The successful application of the
technology promises to mitigate many of the concerns surrounding
handling of these wastes.

CHEMICAL HAZARDS FROM
COMPRESSED GASES
  Many compressed materials exhibit chemical hazards which must
be taken into consideration in managing their disposition. These
hazards include characteristics of ignitability, reactivity, corrosivity.
toxicity and combinations of these characteristics. It is  extremely
important to identify the contents of compressed gas cylinders to
determine the nature of the risks posed by the wastes.
  Many of the gases which are not otherwise hazardous often form
a danger to workers through asphyxiation.  In order to be safe for
human respiration, the atmosphere must contain a minimum of
19% oxygen. Any gas capable of displacing oxygen offers a poten-
tial hazard from asphyxiation. This danger is most common in con-
tained  environments with limited air circulation.
  The  asphyxiation hazard posed by some gases is reduced due
to their good warning properties. For example, any worker entering

-------
an environment which contains high levels of ammonia will im-
mediately recognize its presence. Other gases, how. ever, are more
insidious. Nitrogen is particularly hazardous because of its poor
warning properties and its common use in enclosed areas. Other
gases which tend to  accumulate in low lying areas can displace
oxygen in these areas (i.e  sulfur hexafluoride).
  Other characteristics of compressed gases have more direct
dangers. Many gases can be extremely toxic to humans through
inhalation or contact with the skin or eyes. Some gases can be
hazardous simply because of the low temperature of storage.
  Many of the toxic gases are used in commercial applications.
Examples of these extremely toxic gases include arsine, phosphine
and diborane which are used in the electronics industry. Injuries
and fatalities have resulted from accidents involving these gases.
One example of the  extreme toxicity of some common gases is
arsine whose TLV  is  0.05 ppm. When escaping from pressurized
containers toxic gases can rapidly reach lethal concentrations in
surrounding areas.
  Flammable gases which present both fire and explosion risks are
some of the most  common  gases (i.e., propane  and acetylene).
Flammable gases must be isolated from sources of heat or igni-
tion. Leaking propane has traversed long distances to sources of
ignition with resultant explosions and fire.
  Other gases pose unique hazards based upon their reactivity.
Oxygen, when exposed to oil or grease, can generate an extremely
powerful explosion. Flourine can react with  metals to create dan-
gerous fires. Leaking flourine cylinders and  lines have resulted in
massive fires consuming not only the container but also associated
structures.
  Corrosive gases form another classification of hazardous gases.
Examples include  anhydrous  hydrogen chloride  and hydrogen
flouride.
  Certain specialty gases  have characteristics which make them
dangerous because of their potential for exothermic polymeriza-
tion. Examples are unstabilized hydrogen cyanide and ethylene
oxide.  The handling of a  hydrogen cyanide cylinder exposed to
fire resulted in an explosion in Texas which cracked bank windows
several blocks from  the site of detonation.
  Hazards are often associated with materials stored in compressed
gas containers even though they may not be pressurized. A tragic
example was the accident involving the sampling and recontaineri-
zation of lecture bottles at an industrial facility in the eastern United
States. A lecture bottle containing pentaborane (a non-pressurized
liquid) was  removed from a transfer device by  an unprotected
worker. Even though there would have been no  indication of
residual pressure,  the liquid pentaborane volatilized killing the
worker and injuring others in the vicinity, including rescue workers.
  Because of the variety of gases and materials stored in cylinders,
an extremely cautious approach must be employed when managing
these wastes. This cautious approach especially should be used
where  the contents may not be positively identified.

REGULATORY CONSIDERATIONS
  The management of waste compressed gas cylinders is governed
by federal, state and  local regulations. The most significant regu-
lations are those governing hazardous wastes and hazardous air
pollutants. DOT regulations apply to the interstate movement of
cylinders.
  DOT regulations govern the transportation and refilling of com-
pressed gas cylinders. Under these regulations (found in 49 CFR),
appropriate containers must be used for  each gas or liquid. The
Canadian Transport Commission (CTC) has adopted identical
regulations for shipment. The "Boiler and Pressure Vessel Code"
of the American Society of Mechanical Engineers applies principal-
ly to stationary containers.
  Under DOT regulations, each cylinder  must be stamped to
indicate information concerning the rating and manufacture of the
cylinder. High pressure cylinders have DOT ratings of between 900
and 6000 lb/in.2. Low pressure cylinders range from 240 to 500
    lb/in.2. Acetylene cylinders are included in a special classification
    and the container is filled with a porous material and the gas is
    dissolved in acetone.
      In order to be refilled, cylinders must meet DOT requirements
    for structural integrity.  Periodic hydrostating of the vessel is
    required. Standards have also been developed for corrosion. Con-
    ditions which may make the cylinder unsafe for transport include
    extreme corrosion, bulging or  damage to the vessel or its valve.
      Prior to transportation, a cylinder should be carefully inspected
    to insure it is suitable for safe  shipment. A leaking cylinder can-
    not be legally transported and  should only be handled by a pro-
    fessionally trained worker. Support can be obtained from the owner
    of the cylinder if it is a major supplier.
      RCRA specifically includes  compressed gases in its definition
    of "solid wastes." A waste gas can become a hazardous waste by
    having a listed characteristic (ignitability, reactivity or corrosivity)
    or by specific listing of its components.
      Historically, there has been  a differentiation between  the
    handling of compressed gas cylinders which are owned by gas sup-
    pliers and those owned by a generator. Where a gas supplier rents
    the cylinder to the user,  it can be returned when  it is no longer
    needed. The suppliers treat  any residual gases as part of their
    manufacturing process under operating permits. This  procedure
    has been used to avoid the necessity of treating residual gases as
    hazardous wastes.
      Where title to the cylinder and its contents is transferred to the
    user, the user may be subject to RCRA regulations for disposition
    of residual waste gases. Under  these regulations,  a  container
    approaching atmospheric pressure is considered to be empty and
    need not be handled  as a hazardous waste. Only those cylinders
    with residual pressure must be treated as  a hazardous waste.
      Even  though  the  regulations may  not cover cylinders at
    atmospheric pressure, the contents may still be hazardous. Certain
    hazardous liquids are often stored in compressed gas cylinders and
    may be present even though not pressurized. Some gases may also
    have reacted to form solid or liquid residues with hazardous charac-
    teristics. Improper handling of these materials can result in injury
    or death.
      Waste compressed materials subject to RCRA regulations must
    be disposed of at a permitted facility. Appropriate disposal options
    include thermal destruction or chemical treatment. Non-regulated
    gases can be vented under controlled conditions by experienced
    personnel.
      Very few options presently exist in the United States for disposal
    of compressed gases at permitted facilities. The authors are aware
    of only two permitted chemical treatment facilities for waste gases.
    At least two incinerators will consider disposal of certain com-
    pressed gases. The list of gases  which can be disposed of at these
    facilities does not, however, include  all compressed gases.
      To accept a compressed gas  for disposal at these facilities, the
    contents of the cylinder must be identified and the valve mechan-
    ism must be in operable condition. In a survey of disposal alter-
    natives for the Corps of Engineers,  it was determined that no
    commercial facility will accept  gas cylinders whose contents have
    not been positively identified.
      Because of the limited disposal options,  some generators previ-
    ously have obtained permission for disposal by detonation of the
    cylinders at remote sites. For this operation, shaped charges are
    attached to the cylinder and detonated. This disposal process may
    also include ignition  of fuel surrounding the cylinder to burn or
    decompose flammable gases. This procedure is not effective on
    many hazardous gases, and its  use has been severely curtailed by
    responsible regulatory officials.
      The risks associated with detonation of unknown cylinders as
    a disposal option recently illustrated. Several unidentified gases
    were detonated at a  remote  site. In one case, the charges failed
    to completely sever the cylinder. The resulting explosion rocketed
    portions of the cylinder  over several hundred yards.
      Because of the limited disposal options, cylinders in dilapidat-

VOLATILE ORGANICS MONITORING REMEDIATION     269

-------
ed condition or whose contents cannot safely be determined have
been an extremely difficult problem. The Chemical Control Cor-
poration Superfund site in Elizabeth,  New Jersey, which explod-
ed and burned in 1980, has not yet been completely remediated
because of unidentified cylinders which remain on-site.
  In these cases  the cylinders cannot  be safely sampled through
the valve mechanism because of its condition or the possibility of
failure during sampling.  Sampling options such as cold or  hot
tapping have been dismissed as presenting too many risks given
the pressures and potential chemical characteristics of the contents.
In meeting the problems associated with these cylinders, a system
has been developed to address the particular hazards associated
with these wastes. The system, which  provides for sampling and
recontainerization of the  cylinder contents  in a contained, inert
environment, was demonstrated at a waste cylinder burial site. The
Cylinder Recontainerization  Vessel  (CRV)  proved  to  be  an
appropriate mechanism to overcome the myriad problems posed
by these wastes.

CYLINDER  RECONTAINERIZATION VESSEL
  Earth Resources Consultants, Inc. (ERC) of Orlando, Florida,
has developed a concept for the management of compressed gases
and liquids through the use of a pressure vessel containment system.
The system was designed  for the remote release and recontaineri-
zation of pressurized gases and liquids in a completely contained,
inert environment. At the heart of this system is ERC's patented
Cylinder Recontainerization Vessel (CRV).
  The vessel and system were designed to accommodate the high
pressures and wide variety of gases and liquids available in com-
mon commercial gas cylinders. The CRV is a steel, ASME-rated
pressure vessel measuring approximately two ft in diameter by nine
ft in length. Appurtenances to the vessel are a cylinder clamping
and locking mechanism, a hydraulic drilling system and a vacuum
purge system. The vessel  has a front loading hatch  for insertion
of the target cylinder. The system incorporates roller mechanisms
for rotating cylinders inside of the vessel.
  Interior pressures and  temperatures are monitored by  remote
sensing units. A  pressure  illumination and remote video viewing
system provide direct observation of the cylinder while inside the
vessel during processing operations. The CRV is located inside a
sealed tractor trailer van. All of its systems are hydraulically or
pneumatically controlled.
  The entire system is operated from a remote panel located out-
side the van. From  the command control panel, operations  can
be monitored and controlled while all personnel are removed from
the process area.
  The CRV incorporates an inert gas purging system. In conjunc-
tion with  vacuum  pumps capable or producing  a  104 Ton-
vacuum, the purge system provides complete removal of released
gases. Diaphragm compressors complete the purging and recon-
tainerization  process.
  A secondary containment chamber houses the CRV and its recon-
tainerization systems. This reinforced steel chamber can be sealed
to contain any release from the primary system. All of the equip-
ment in the chamber is suitable for operation in a Class I. Divi-
sion I explosive environment. Parts are attached so that any released
gases can be withdrawn and treated in the unlikely instance of a
leak. Additional  structural support  is  provided by the reinforced
semi-van trailer housing the unit. This  trailer can be located inside
a containment tent also covering the  working area.
  A major reason for using this system is to obtain samples for
identification  of cylinder contents. The  CRV incorporates sampling
ports for gas withdrawal. A sample is obtained and transferred
directly to an on.site analytical unit for analysis. The analysis is
completed in a matter of minutes, and the released gas is then recon-
tainerized  into a DOT-approved vessel.
  The CRV was developed for use on  cylinders whose valves are
no longer operable or which are in dilapidated condition.  Corro-
sion associated with weathering of cylinders often produces these

270    VOLATILE OROANICS MONITORING REMEDIATION
circumstances. Burial of gas cylinders for extended periods is likely
to degrade the cylinder. The CRV was first used at a gas cylinder
burial site in Jeffersonville, Indiana.

REMEDIAL ACTION AT WASTE GAS
CYLINDER BURIAL SITE
   The U.S. Army Corps of Engineers (COE) was responsible for
implementation of a remedial action at the Jeffersonville Quarter-
master Depot in Jeffersonville, Indiana. As part of its function
in the manufacture of uniforms, the Depot was responsible for
the fumigation of  clothing worn by  military personnel.  One
mechanism used to delouse the clothes was fumigation with methyl
bromide. Between 1952 and 1954, canisters of methyl bromide were
buried in a designated disposal area in the northwest corner of the
Depot. Historical research by COE investigators indicated that ap-
proximately 100 canisters were buried in a trench 200 ft long by
30 ft wide by 6 ft deep. The site was fenced and placarded. Fol-
lowing deactivation of the Depot, the property was sold and cur-
rently is used as a commercial warehouse area.
   During  1986, the COE contracted with  ERC to excavate the
cylinders,  recontainerize the contents and dispose of the waste. On-
site activities were initiated in December of 1986 and concluded
during January of 1987.
   Methyl  bromide, also known as bromo methane, is a colorless
liquid or gas with a chloroform-like odor only in extremely high
concentrations. Methyl bromide causes no immediate nose or
respiratory irritation even in poisonous concentrations. These poor
warning properties  make  handling methyl bromide hazardous
because no exposure signal is given until high concentrations have
been encountered. Methyl bromide gas is a severe pulmonary irri-
tant and neurotoxin. It is also a narcotic at  high concentrations.
Routes of exposure include inhalation, dermal or eye contact and
oral intake.  The current OSHA standard for  methyl bromide is
a ceiling of  20 ppm.
  The methyl bromide canisters presented a serious hazard to both
the site workers and the surrounding area. In  developing a tech-
nical approach to the project,  prime consideration was given to
worker safety, protection of surrounding businesses and residences
and protection of the environment. The  project included an
extensive health and safety program, a  public relations program
and engineering systems to control on-site operations.
  The primary hazard posed by the project was a potential release
of a large quantity of concentrated methyl bromide, leading to
respiratory inhalation or dermal exposure.  Workers involved in
excavating and handling these materials  required both respiratory
and dermal protection, which  was afforded by the protective
clothing chosen for this cleanup.
  While protection of site personnel was provided in the safety
precautions, consideration also was given to off-site safety of the
local community. Because of the close proximity of the site to both
businesses and residences,  precautions were taken to assure that
adequate protection was provided in the case of an uncontrolled
release of the methyl bromide gas.  Special engineering  and
monitoring controls included the use of a containment structure
over the area of excavation, an air treatment system, air monitoring
and computer programs modeling potential dispersion conditions
and rates.
   A containment structure was necessary because of the location
of  warehouses within 100 ft of  the excavation and the close
proximity of residential neighborhoods, making removal of toxic
gas containers a threat to both on and  off .site personnel. Based
on air modeling programs, it was determined that a release of one
cylinder of methyl bromide could result in potentially hazardous
levels of the gas reaching either the warehouses or nearby residences
in a matter of seconds. Based on this assessment, there would be
insufficient  time to evacuate these areas should a release occur.
To mitigate this hazard, a large sealed containment tent measuring
approximately 100 ft long by 30 ft wide by 14 ft high was erected
over the burial area. This sealed structure was provided with air

-------
lock doors for entry and exit.
  As a second precaution, an air treatment system was provided
for filtering the air inside the building in the event of a release.
The system was sized so that the entire atmosphere inside the
building could be processed through the treatment unit within
approximately one hour.
  Given the regulations governing transportation of compressed
gas cylinders and the suspected condition of these cylinders, the
selected approach was to remove the contents from the cylinders
and transfer them into new DOT-approved vessels for off. site treat-
ment. Since the valves would very likely be corroded and unrelia-
ble for gas transfer, ERC proposed to use the CRV to accomplish
the transfer.
  Since available information indicated that the cylinders  were
buried approximately six ft deep, it was determined that the primary
excavation would proceed using a tractor backhoe to a depth of
four ft. Workers would then enter the excavation and continue
exhumation by hand to the depth of the cylinders.
  Hand excavation techniques were tailored to provide careful
cylinder extraction. Soil was removed in 1 in. to 2 in. lifts  until
the cylinders were encountered. This method minimized the poten-
tial for rupturing deteriorated cylinders.
  The initial cylinder was discovered within the first few hours of
excavation. The cylinder was exposed using hand tools,  and was
examined in the excavation for overall integrity before removal.
Visual inspection revealed the cylinder to be in fair condition with
no  apparent leaks.
  Six cylinders were removed by the  above procedure during the
first day of operations. Three of the cylinders were 9 in. in diameter
by 54 in. in length, with an inverted disc bottom and foot ring typi-
cal of the low pressure canisters used to contain methyl bromide.
All of the cylinders had protective caps over the valves. Although
protected, the  valve  mechanisms appeared to  be  completely
inoperable.
  Also uncovered during the initial excavation was a cylinder of
high pressure design. There were no identifying labels as to its con-
tents. The valve  found on the  cylinder  was a type typically
associated with non-flammable, non-poisonous gases. Although
the cylinder was intact, the valve was extremely corroded and
appeared  to be inoperable.
  During this excavation, small glass ampules were uncovered in
the upper 2 ft of the trench. The excavation was immediately halted
to determine the nature of this material. Several ampules were col-
lected from the pile, each containing approximately 75 ml of clear
or yellow liquid.
  When first discovered, the OVA readings indicated an increase
in organic vapors inside  the containment structure. The fragments
were assumed to be from broken  containers. With the  elevated
OVA readings, the air treatment system was immediately started
and all vents to the outside were closed. Organic vapor readings
indicated  concentrations of up to 75 ppm total organic vapors
within the containment structure.
  Within  1 hour, OVA readings had  fallen to background levels,
indicating  that the carbon scrubbing system  was  successfully
removing  the methyl bromide vapors.  Charcoal sampling tubes
from the air sampling pumps were rush delivered for laboratory
analysis. These revealed that concentrations were below the OSHA
TLV-TWA of 5 ppm.
  Recontainerization of the full methyl bromide cylinders pro-
ceeded in  accordance with a plan submitted to the COE prior to
site activities. All the CRV systems were pre.tested and determined
to be fully operational. Drilling of cylinders commenced only after
testing was complete.
  The actual drilling took only several seconds per cylinder. Video
observation of the canisters indicated very little gaseous release
upon penetrating the cylinders, which was expected because the
methyl bromide was maintained in its liquid phase at 30 ฐF by a
refrigeration unit mounted in the trailer. No increase in pressure
was noted inside  the vessel during the operation.
  After drilling the first hole, the cylinder was rotated 180 degrees
using the rollover system. Once fixed into position, a second hole
was drilled into the canister. This hole facilitated drainage of the
liquid content.
  After completion of the second hole, the CRV was hydraulically
tilted  to facilitate drainage of the liquid by locating the liquid
removal drain at the lowest portion of the vessel. At this point,
the vessel was slightly pressurized with nitrogen through purge gas
vents, forcing liquids  from the vessel. Once the vessel was pres-
surized, valves were remotely opened which permitted the liquid
to move through a piping system, through filters and into the new
containers. This process was repeated several times until all of the
contents had been transferred from the  vessel into new cylinders.
  At  the conclusion  of the methyl bromide transfer, the entire
system was purged with a Freon liquid. The purge liquid flushed
the system and was transferred into one of the containers for trans-
port to the disposal facility. Purge gas insured that the interior of
the cylinder and its lines and filters  were cleared of the methyl
bromide liquid at the conclusion of  the operation.
  Once the recontainerization effort was complete, the remaining
high pressure cylinder was addressed. Although this cylinder could
not be positively identified, it was believed that it most likely con-
tained carbon dioxide. This conclusion  was based on identifying
marks located on the valve and conversations with the manufacturer
of this valve.
  Processing this cylinder indicated that the contents were indeed
a pressurized  gas. Immediately following  the rupture of this
canister, pressure inside the containment vessel rose to approxi-
mately 38 lb/in.2. A sample was withdrawn from the sampling
port and identified on.site as carbon  dioxide. A sample also was
obtained for laboratory confirmation.
  All the materials recovered from the excavation were disposed
of as hazardous waste. The empty methyl bromide cylinders, cans
and glass  vials were  overpacked and  disposed  of  at  a secure
hazardous waste landfill. The recontainerized methyl bromide was
transported to a permitted treatment  facility. At this facility, the
methyl bromide was released into a caustic scrubbing system which
resulted in generation  of inert salts and water. Contaminated car-
bon from  the air  treatment system  was  transported to  an
incinerator.
  All of the objectives of the project were met. The methyl bro-
mide  cylinders were successfully recovered and recontainerized.
Because of the condition of the cylinders, they could not have been
safely transported off.site without the recontainerization opera-
tion.  Given the condition  of the valves  on the cylinders, recon-
tainerization by other  means would not have been safely feasible.
  The health and safety plan, the project specific engineered con-
trols and the tailored air monitoring plan associated with the project
also were successful. The air monitoring program indicated no sig-
nificant release of methyl bromide vapors outside the containment
tent or CRV van trailer. No workers  were exposed to the poten-
tially  toxic methyl bromide vapors. The most evident success  of
these  safety controls was the successful operation  of the air treat-
ment  system during the unexpected release of vapors associated
with the discovery of the glass ampules.
  Finally, the most significant project success was associated with
the recontainerization of the high pressure gas cylinder. This project
represented the first time that an unknown gas cylinder, in deteri-
orated condition, was safely sampled in a totally controlled manner.
The CRV system also permitted the release of compressed liquified
gas and recapture in stable, transportable containers. Thus, the
success  of the CRV was twofold: sampling and identification of
an unknown compound and the transfer of known toxic com-
pounds into safe containers.

CONCLUSION
  The CRV has application at other sites where gas cylinders may
contain unidentified gases or hold known hazardous compound
in deteriorated containers. The construction and operation of the
                                                               VOLATILE ORGANICS MONITORING REMEDIATION     271

-------
CRV contains many engineering controls for the safe handling of
a wide variety of gases. These controls include the following:

• An ASME rated pressure vessel capable of accommodating the
  highest pressures contained in manufactured gas cylinders
• Use of inert passivated materials such as steel, stainless steel and
  Teflon to minimize potential reactions
• Capability for removal of the air inside of the vessel prior to
  processing operations
• Replacement of the air with an inert gas atmosphere prior to
  drilling operations
• Use of nonreactive hydraulic systems inside the sealed van trailer
  to avoid ignition of  flammable or explosive gases
• A recontainerization system to provide for the transfer of the
  contents of the released gas in a contained loop
• Capabilities for on-site withdrawal and analysis of the released
  gas prior to recontainerization
• Complete system purge to prevent oxidation reactions, (e.g., oxy-
  gen with residual hydrocarbons
• Remote operation of the  unit by personnel outside of the CRV
  van. This control process includes the capability to monitor the
  pressure and temperature conditions during operation as well
  as visually inspect through a remote viewing system
• Location and operation inside a sealed mobile van trailer which
  can withstand pressures of up to 10 Ib/in. This sealed trailer has
  withdrawal ports so that, in the event of a leak from the CRV
  system, the released gases can be processed through an on-site
  carbon absorption  and caustic scrubbing system
• An effective plan for personnel protection including the use of
  totally encapsulating suits with a supplied air system
• These engineering controls incorporated into the CRV provide
  the only  viable, operational system  to sample and recon-
  ainerize compressed gas cylinders.
  The Jeffersonville project combined traditional remedial tech-
niques (excavation) with new technology to locate, remove and
transfer the cylinder contents. It also demonstrated the ability to
employ complementary engineering devices to mitigate the poten-
tial for exposure to a dangerous compound.
  The management of compressed gas cylinders in deteriorated
condition requires experienced, professional services because  of
the variety of potential risks. Although risks are not as severe, gas
cylinders in good condition still deserve considerable respect and
handling by trained personnel.  The possible regulation of these
materials under hazardous waste regulations requires that these
materials  be accorded special  attention in waste management
programs.
272     VOLATILE OROAN1CS MONITORING REMEDIATION

-------
             Vacuum  Extraction  of  Hydrocarbons  from  Subsurface
                           Soils  at  a  Gasoline Contamination  Site

                                                   Joseph Applegate
                                                John K.  Gentry, P. E.
                                Florida Department Of Environmental Regulation
                                                 Tallahassee, Florida
                                                 James J. Malot, P.E.
                                               Terra Vac,  Incorporated
                                                 San  Juan Puerto Rico
ABSTRACT
  Terra Vac, Inc., in cooperation with the Florida Department of
Environment Regulation (FDER), installed and tested a vacuum
extraction system at a gasoline contamination site in Belleview,
Florida where a substantial leak of unleaded gasoline had contami-
nated subsurface soils and groundwater. With the increasing
number of leaking underground petroleum tanks in Florida, it is
essential to demonstrate new cleanup technologies that are efficient
and cost-effective. The commonly used cleanup methodology con-
sisting  of free product  recovery and groundwater withdrawal
without soil remediation is very costly and lengthy since the con-
taminated soils serve as a continuing  source  of contaminants
leaching into the groundwater.
  The  site  came to the FDER's  attention  in  1982 when the
municipal well field, which is developed within the Floridan aquifer,
became contaminated with gasoline components and had to be
abandoned. A subsequent investigation revealed that at least 10,000
gal of unleaded gasoline had leaked  between October 1979 and
March  1980 from a service station located 600 ft up-gradient from
the wellfield.
  In this region of Florida, the Floridan aquifer, which is the
primary source of groundwater, is  comprised of a highly karstic
limestone in which dissolution  has formed an extremely porous
and permeable aquifer system.  Due to the discontinuous nature
of the overlying clay layer in the vicinity of the site, the Floridan
aquifer is under water  table conditions. Thus, there is  little
resistance to hydrocarbons migrating into the aquifer from  over-
lying soils containing residual hydrocarbons.
  Because of the 50-ft thickness of the unsaturated zone and the
large quantities of contaminants in the subsurface soils at the site,
FDER sponsored a pilot study to demonstrate the effectiveness of
vacuum extraction technology  to treat soil contamination. The
Terra Vac system is designed to recover both free product and
adsorbed hydrocarbons, in situ, from the contaminated soil.
  The objectives of the pilot study were to: (1) delineate the extent
and magnitude of hydrocarbons in the subsurface; (2) quantify the
rate that hydrocarbons can be extracted; and (3) develop a con-
ceptual design and time-frames for the cleanup of soils at the site.
  The preliminary results are excellent. The Terra Vac  Vacuum
Extraction System recovered more than 8000 Ib of hydrocarbons
from the subsoil during the three week pilot test at rates of  up to
2000 Ib/day. The average recovery rate during this period was about
880 Ib/day. FDER has since extended the project to further  clean
up the  site.
  This  unique extraction process solves many problems that are
inherent to other cleanup alternatives by first removing the VOCs
    from the vadose zone and then follow with groundwater recovery
    and treatment, if required. For VOC soil contamination, the Terra
    Vac Vacuum Extraction System is cost effective and is a "total"
    approach to decontamination.

    INTRODUCTION
      The Florida Department of Environmental Regulation (FDER)
    is responsible for conducting the assessment and remediation of
    certain sites that are contaminated with petroleum hydrocarbons
    and other non-hazardous or hazardous materials. When a site
    becomes eligible for State cleanup, a Project Manager is assigned
    and contractors of the Department are employed to perform the
    tasks required to complete the cleanup. As part of the remedia-
    tion process  for the Belleview gasoline contamination site, the
    Department contracted Terra Vac, Inc., to install their vacuum
    extraction system and demonstrate its effectiveness to reduce con-
    taminant levels in  the subsurface soils, the source of on-going
    groundwater contamination.
      The inventory of sites in the State of Florida that are contami-
    nated with petroleum hydrocarbons presently has more than 1000
    listings. As of June 1, 1987, petroleum contamination at 1,687 sites
    bad been reported under the State's Early Detection Incentive
    program. The high incidence of petroleum contamination from
    leaking underground storage systems and spills requires that an
    efficient and cost-effective solution be found to achieve a cleanup
    of these sites.
      The Department's experience with the remediation of petroleum
    hydrocarbons contamination has shown that when soils are con-
    taminated and allowed to remain in place untreated, the cleanup
    process is costly, lengthy, and may never reduce contaminant levels
    enough to use the water resource as a potable supply again.
      The method most commonly used to remediate soil contamina-
    tion is removal. Transporting the  removed  soils to a suitable
    disposal location is expensive and merely transfers the contami-
    nation, rather than eliminating it. Removal is also disruptive and
    may be impractical due to physical conditions at the site, such as
    the locations of buildings,  roadways or utilities. The FDER  is
    seeking,  through the used  of its contractors, to evaluate and
    optimize treatment technologies for soil contamination. This paper
    details conditions at the Belleview site and presents the results of
    the in situ soil treatment pilot study.

    SITE HISTORY
      The Belleview gasoline contamination site came to the Depart-
    ment's attention in late 1982 when residents complained that water
    from the municipal well field had an objectionable taste and odor.

VOLATILE ORGANICS MONITORING REMEDIATION     273

-------
Analyses by the FDER Span Lab of water  samples from the
municipal wells  detected the following  total  concentrations of
benzene, ethylbenzene, toluene and xylenes (BTEX);

City Well #\  =  66 ug/1
City Well #2 =  15 ug/1
City Well #3 =  470  ug/1

  Based on the compounds detected and the pattern of the chro-
matograms, the Department concluded that the  contamination was
a partially weathered  gasoline1. Since  the wells all contained
benzene in excess  of the Department's Maximum Contaminant
Level (MCL) of 1 ug/1 for the compound, all of the wells  had to
be abandoned. A subsequent inspection of the inventory records
of gasoline stations in the area revealed that between October 1979
and March 1980, at least 10,000 gal of unleaded gasoline had leaked
from a service station located approximately 600 ft upgradient from
the city wellfield.
  An inspection of the station by the State Fire Marshall found
explosimeter readings of more than 100%  under the concrete apron
on  the  east side of  the station. Based  on these findings, four
monitors wells were  installed in the area. MW-2, located nearest
to and downgradient of the station, contained  0.25 in of gasoline
free product. Analysis of the groundwater from well MW, 2 found
33,090 ug/1 of benzene, ethylbenzene, toluene, and total xylenes
(BTEX).
  Because the station owners were financially unable to undertake
a cleanup of the leakage, the site became eligible for State cleanup.
The first step in  that process was an initial assessment of the site
completed by the Department contractor  in 1985. This assessment
                                        464
                                  C*ndlซr~\ .MOMBluff
                                          OUiwtht
                          Figure 1
      Location of Union 76 Service Station Belleview, Florida

274    VOLATILE OROANICS MONITORING REMEDIATION
defined conditions at the site and verified that large quantities of
petroleum hydrocarbons existed in the subsurface2. The results of
the contamination  assessment  were the basis  for the vacuum
extraction pilot study at  the site.
SITE CONDITIONS
  The Belleview  gasoline contamination  site  is  located at the
intersection of Robinson  Road and U.S. Highway 441 in Marion
County in the North Central portion of Florida. Its location is
shown on Figure 1. Land surface elevations are approximately
100 ft above mean sea level. Two surface water bodies. Lake Lillian
and the Grotto, are located near the site. Lake Lillian is a relic
sinkhole in the karst topography; the Grotto is still active, having
a hydraulic connection with the Floridan aquifer.
  The surficial geology in the site vicinity consists of three gener-
alized units: (1) a surficial unit consisting of pleistocene sands, silts
and clayey sands (2) an intermediate semi-permeable confining unit
consisting of Miocene clays and (3) sands and sandy clays. The
underlying Eocene and  older  limestones  which  compose the
Floridan aquifer,  is the sole source of groundwater. This  latter
formation is compromised of a highly karslic limestone in which
dissolution has formed an extremely porous and permeable aquifer
system. Monitor wells drilled into the limestone have encountered
voids and large sand filled cavities. The Belleview wellfield is
developed into the Floridan aquifer. Well DW-3, the most heavily
contaminated municipal  well, is typical of those wells and is
198 ft in total depth. It contains 88 ft of casing. Lithologic condi-
tions in the area around the site are highly variable but generally
conform to the geologic  description given above.
  The surficial  sands contain very little water. Where a shallow
water table is encountered, it exists as a perched zone of limited
area I extent situated above relatively impermeable clays at depths
of about 20 ft. The clays beneath the site are inter bedded with sand,
causing them to  exhibit  relatively high permeabilities in  some
locations. Tests performed by the FDER contractor on Shelby tube
samples of the clays determined their permeability to range from
2.45 x  ID'3  to 7.81 x  1Q-* cm/sec, averaging I IQ-* cm/sec.
The variation in the composition of the confining unit and the
limited areal extent  of the more competent clays produce varying
degrees of confinement on the underlying Floridan aquifer. The
relatively permeable nature of the confining unit and the changing
elevation of the top of the limestone cause conditions where the
aquifer is confined and under artesian conditions in some portion
of the area, while in others, it exist under water table conditions.
Contaminants from the surface have access to the Floridan aquifer
through the more permeable zones in the confining unit.
  Regional potentiometric surface maps of  the upper Floridian
aquifer  indicate that the downgradient direction of groundwater
flow in the aquifer is northwest, directly from the gas station site
toward the municipal well field.  Water level  measurements from
the monitor wells  at the site have shown the downgradient direc-
tion to vary from west to northwest to northeast. The most recent
water level contours (April 1987) indicate the groundwater gradient
is in the northwest direction. The Belleview municipal wells have
not been operational since they were taken out of service in  1982.
When they were pumping, they no doubt exerted considerable in-
fluence on the flow regime and controlled the hydraulic gradient
within their radius of influence which extended to at least the site
of the gasoline leak.
  The assessment work performed  in 1984 determined that there
were still components of  gasoline in the Floridan aquifer down-
gradient of the station. It  also verified the existence of a relatively
thick unsaturated  zone, at least for the state of Florida, averaging
about 50 ft. The assessment recommended that the next phase of
the site investigation should determine the presence of and quantify
the amount of gasoline in the soil at the station. Soil borings should
be drilled in the area of the leak to determine if organics are trapped
near the source of the leak. The feasibility of removing the source
of groundwater contamination could then be determined.2

-------
THE PILOT STUDY
  The FDER decided to evaluate the effectiveness of the Vacuum
Extraction Process for the treatment of soil contamination. This
vacuum extraction system is designed to recover both free product
and adsorbed hydrocarbons from the contaminated soils without
excavation. Because of the magnitude of the leak and the thick-
ness of the unsaturated zone, Belleview seemed  an ideal site for
a pilot study.

The objectives of the pilot study were to:

• Delineate the extent and magnitude of hydrocarbons in the sub-
  surface
• Quantify the rate at  which hydrocarbons can be extracted
• Develop a conceptual design and time-frame for the cleanup the
  site

  Work began in December 1986 when four soil  borings were
drilled at locations shown in Figure 2. Soil samples were obtained
from each boring to assess the subsurface conditions and quantify
the hydrocarbons present. Soils were classified by a geologist and
analyzed on-site for gasoline components.  Benzene, toluene and
xylenes  and total hydrocarbons were quantified in soil samples by
gas chromatography using the headspace method. Selected samples
were sent to a certified laboratory for correlation purposes. The
results of the soil testing are shown on Table 1.
                           Table 1
         Soil Concentrations By On-Site Headspace Analysis
                        DEPTH
                        (FT)
SOIL CONCENTRATION tPPBJ
BENZENE  TOLUENE  XYLENES
                          Figure 2
                     Well Location Map
   Soil samples were obtained from three locations in the tank pit
 area. A hand auger was used to collect samples for on-site hydrocar-
 bon analysis. Selected samples were  set to the  laboratory  for
 analysis. Results of these analysis are shown in Table 2.
   A vacuum extraction/monitoring well was installed at each of
 the test boring locations. Two wells, VE-1 and VE-2, were installed
 primarily for vacuum extraction of subsurface hydrocarbons from
 above and below the confining unit, respectively. At the other two
 boring locations, multi-level, dual purpose wells were installed.
 These  wells were designed to monitor subsurface  vacuum and
 extraction of hydrocarbons from three different hydrogeologic
 zones.
   In boring ME-1 three vacuum monitoring wells were installed:
 ME 1-13, ME-1-35 and ME-1-50 monitored subsoils near  13, 35
 and 50 ft deep, respectively. ME-1-13 and ME-1-50 are also capable
 of vacuum extraction of hydrocarbons from these zones above the
 clay layer and separately within the limestone formation. Similarly,
 in the borehole at ME-Z, two monitoring/extraction wells, ME-2-16
 and ME-2-58, were installed to monitor and extract hydrocarbons
 from clayey soils approximately 16 ft deep and from the limestone
EXTRACTION WELL
VE-1
1
2
3
4
5
6
7
a
9
10
11
12
13
14
IS
16
17
19
19
EXTRACTION WELL
VE-2
1
2
6
3
S
6
7
8
9
10
11
12
13
14
IS
16
17
18
19
EXTRACTION WELL
HE-1
1
2
4
3
S
6
7
a
9
10
11
12
13
11
IS
16
17
EXTRACTION WELL
ME-2
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
18
19


1
3
6
10
12
15
17
18
21
23
26
28
31
33
35
37
40
44
46


3
7
12
10
15
17
20
22
25
27
35
44
40
46
49
52
54
S7
60


3
7
12
9
15
17
20
22
25
27
30
32
35
37
40
42
46


3
7
10
12
15
17
20
22
25
27
30
32
35
37
40
42
47
50


N/A
232
10.900
1,877
29,345
730
526
1.160
466
270
113
70
28
60
26
43
75
175
166


2
4
HO
4
304
2.744
877
708
230
93
36
NO
14
2
1,216
3,367
8
NO
ND


158
1.476
19,500
2,350
7.324
6.582
439
360
152
190
92
10
11
ND
NO
ND
2, 410


201
906
8,873
2,065
2,184
825
155
500
538
229
112
61
176
107
55
15
237
925


1,103
606
28,180
4,264
38,518
1.701
596
1.140
518
205
92
54
28
47
40
53
103
428
470


4
3
ND
2
98
3.266
121
9
BO
33
24
ND
4
3
1,335
5,788
77
21
ND


35
1,657
24,200
3,397
10,871
11,649
361
175
76
165
125
IB
24
ND
ND
. 22
7,170


53
1 867
18 971
4 739
4 549
\ 1BO
162
1.306
699
171
45
72
281
213
95
30
1.063
2.860


2.486
511
16,730
1,586
29,068
4,597
252
1,090
686
256
122
114
22
107
65
55
136
S5B
527


ND
1
113
2
72
4.035
33
9
37
103
25
2
ND
2
5,692
15.784
360
NO
ND


ND
480
6,980
1,084
3.867
8,111
253
106
118
197
95
0
0
ND
ND
ND
10,900


63
1,209
17,910
3,089
3,242
1,100
105
77S
591
353
72
78
269
217
140
55
1 ,499
1,780
                 BCLOU DETECTABLE LIMIT OF 1 PPB
at a 58-ft depth. A profile of these extraction wells and the sub-
surface stratigraphy is shown in Figure 3.
  Each well was connected to the vacuum recovery unit through
a manifold system. Since the gas station is an active, operating
service station, special considerations were provided so that the
installation and operation of the vacuum extraction system did not
interfere with the operations of the service station. Well comple-
tions  and the  vacuum extraction manifold were installed using
underground valve boxes and were covered with concrete.
                                                               VOLATILE ORGANICS MONITORING REMEDIATION    275

-------
                           Table 2
     Soil Concentrations Determined by Off-Site Laboratory Test
      SAMPLE
                     DEPTH
                     (FT)
                              SOIL  CONCENTRATION  (PPB)
                              BENZENE  TOLUENE  XYLENES
EXTRACTION WELL
      VE-1

       3
       8
      10

EXTRACTION WELL
     VE-2
      17
      18

EXTRACTION WELL
    ME-1
       8
      17

EXTRACTION WELL
    ME-2

       6
      14
      19
W
                       6
                      18
                      23
                      12
                      54
                      57
                      12
                      22
                      46
                      17
                      37
                      50
 6,270    97,000   210,000
   157       160       <134
  <103      <103       <103
                                <107
                                           < 112
                                           <107
                        .112
                        <107
                        ( 115
12,000    69,000    127,000
   <91       <91        <91
 1,170    21,700     98,600
 765     1.050
<107      <107
<103       310
                       760
                      < 107
                       233
                            Figure 3
                       Subsurface Profile
  On Jan. 29, 1987, vacuum extraction of the subsurface hydro-
carbons began. Initially, vacuum extraction of VE-1 and VE-2 were
conducted both  separately  and together, while the subsurface
vacuum was measured in  five monitoring  wells. During  the
following week, 1-day vacuum extraction tests were conducted on
each  well  that was designed for vacuum extraction. Relative
recovery rates of hydrocarbons from pilot test wells ranged from
450 to 2000 Ib/day. Six wells (VE-1, VE-Z. ME-1-13, ME, 1-50,
ME-2-16 and ME-2-58) can be used simultaneously or individually
for extraction of hydrocarbons.  Quantification of subsurface
vacuum, hydrocarbon recovery rates and overall hydrocarbon con-
centrations were monitored on a daily basis.
  On Feb. 10, unusually high levels of hydrocarbons were being
extracted from the system and liquid product was observed in the
manhole at well ME-1. At the request of representatives of FDER,
the underground storage tanks and product lines were tested.
Results indicated that the product lines  failed the test. Further
investigation revealed a corrosion hole in the pipeline as the source
                                 of additional subsurface hydrocarbons. After the leak was repairld,
                                 vacuum extraction was resumed. The occurrence of the leak was
                                 in no way related to the vacuum extraction operations. Rather,
                                 it was fortuitous that the vacuum system was in place at the time
                                 of the leak so that cleanup was initiated immediately and any migra-
                                 tion of hydrocarbons was minimized.
                                   Subsequently, all the wells were extracted simultaneously to yield
                                 the maximum recovery rate, up to 2000 Ib/day. Recovery rates were
                                 monitored on a daily basis except for maintenance. The system
                                 was modified to include an in-line air/water separator to separate
                                 small amounts of water and product that were being extracted with
                                 the  hydrocarbon vapors.4

                                 RESULTS
                                 Soil samples were analyzed from each of the four vacuum extrac-
                                 tion well locations and three shallow borings in the tank pit area.
                                 The extent and magnitude of the subsurface contamination in these
                                 locations are summarized in Figure 4 for benzene. Similar treodi
                                 are apparent for both indicator parameters and total hydrocarbon!.
                                w                                                         t
                                                                                            Figure 4
                                                                                        Benzene Contours
                                                                   The highest levels of benzene and total hydrocarbons up to
                                                                 29,000 ppb benzene and 335,000 ppb THC, were observed beneath
                                                                 the product pipeline, the suspected source of the subsurface con-
                                                                 tamination. The maximum level of benzene observed in the tank
                                                                 pit was  1,200 ppb at a depth of 10 ft.
                                                                   Concentrations generally decreased with depth and distance from
                                                                 the leaky pipeline until the perched groundwater was encountered
                                                                 just above the clay layer about 17 to 20 ft deep. Above the day
                                                                 layer, benzene concentrations ranged from 1,100 to 6,600 ppb and
                                                                 extended to VE-2, approximately 50 ft away from the source.
                                                                 Similarly, total hydrocarbon concentrations were about the same
                                                                 beneath the leaky pipeline as those observed more than SO ft away
                                                                 at VE-2.
                                                                   Apparently, the clay layer at 17 to 20 ft has caused significant
                                                                 lateral migration of hydrocarbons from the source area beneath
                                                                 the leaky pipeline.  Based on the magnitude of the original leak,
                                                                 an estimated  10,000 gal and the free product observed in down-
                                                                 gradient groundwater monitoring wells, free product apparently
                                                                 migrated laterally across the top of the clay layer over much of
                                                                 the site  where the clay exhibits low permeability. Since the clay
                                                                 layer is discontinuous at the monitoring well locations beyond the
                                                                 site  boundaries, significant  amounts of residual  hydrocarbons
                                                                 remain in the soils above clay as a lingering source of groundwater
                                                                 contamination. Determination of the extent of soil contamination
                                                                 above the clay is necessary to effectively design and implement a
                                                                 system to cleanup the site.
276    VOLATILE ORGANICS MONITORING REMEDIATION

-------
   Beneath the clay layer, at depths of approximately 30 ft, con-
 taminant concentrations are about 10 times lower within the silty
 sand strata and the  upper portion of the weathered limestone.
 However, limestone samples near the water table, approximately
 50 ft deep, exhibited relatively high levels of benzene and total
 hydrocarbons, up to 3400 and 95,000 ppb, respectively. Even after
 several years of natural flushing of the weathered limestone by the
 rapidly moving  groundwater flow, a "smear zone" of residual
 hydrocarbons near the water table still persists as an immediate
 threat to groundwater quality.
   Comparison between analysis of hydrocarbons by gas chroma-
 tography (GC) on-site and off-site laboratories can be made from
 Tables 1 and 2. In general, the on-site GC using the headspace
 method is about 100 times more sensitive than the off-site labora-
 tory which used a methanol extraction method. At very high soil
 concentrations, greater than about 500 ppm, the headspace of the
 sample container approaches vapor saturation and the laboratory
 methanol extraction method yields higher  results. However, at
 hydrocarbon concentrations less than about 10 ppm, the labora-
 tory method yields lower values or the chemicals are not detecta-
 ble. Apparently, significant losses of volatile contaminants occur
 between the time of sampling, preservation, transportation, labora-
 tory preparation and analysis, even though standard protocols for
 sampling and analysis were observed.

 VACUUM EXTRACTION PROCESS
   Implementation of the vacuum extraction process was first con-
 ducted on individual wells to determine initial recovery rates.
 Hydrocarbon recovery rates from vacuum extraction  wells are
 shown in Figure 5. The highest extraction rates were observed in
 wells corresponding to the highest soil concentrations. These data
 suggest a  correlation between soil concentrations and  vacuum
 extraction rates. Although lower  flow rates were observed in the
 soils above the clay layer, relatively high concentrations from these
 wells yielded recovery rates between 300 and 950 Ib/day. Con-
 versely, the concentration of hydrocarbon vapors in the limestone
 was relatively low. However, large flow rates obtained from VE-2
 and ME-1-50 yielded recovery rates  of  1650 and 2000 Ib/day,
 respectively.
    2500
  10
  CD
    1500-
  g 1000-
                    VE-2   ME-1-13  UE-1-50 ME-2-16  UE-2-58
                          EXTRACTION WELL


                          Figure 5
                   Relative Extraction Rates
  This high recovery rate from just above the water table suggests
that pockets of liquid product were hydraulically trapped within
the limestone, yet effectively recovered by the vacuum extraction
process.
  About 2 weeks after the start of the vacuum extraction pilot test
recovery rates increased dramatically in response to another leak
in the pipeline. Recovery rates of up to 1350 Ib/day were observed
 from ME-1-13. After the leak was repaired it took about 1 month
 of vacuum extraction to reduced vapor concentrations to levels
 observed prior to the pipeline  leak.
   During the pilot test and subsequent vacuum extraction opera-
 tions about 22,000 Ib, of hydrocarbons were extracted from the
 subsoils at the site. As the subsoils are cleaned up by the vacuum
 process, the extracted vapor concentration and the hydrocarbon
 recovery rates decline with time. Figure 6 illustrates hydrocarbon
 extraction rates  at the end  of the demonstration project.  The
 reduction in extraction rates is due to the decline in concentrations
 that were observed at the site  as the cleanup progressed. Other
 vacuum extraction wells indicated similar results. In addition, sig-
 nificant reductions in dissolved  hydrocarbons in  downgradient
 monitoring wells have been  observed as a result of the vacuum
 extraction process.

                            AUGUST 1987
!^ 2-
         VE-1      VE-2     ME-1-13  UE-1-50
                          EXTRACTION  WELL
                                            ME-2-16   ME-2-53
                            Figure 6
                Relative Extraction Rates August 1987


   Comparison of the initial extracted vapor concentrations and
 the  final vapor concentration is useful to evaluate the relative
 cleanup level that was achieved by the vacuum extraction process.
 Figure 7 indicates the relative cleanup achieved at VE-1 during the
 project. Similar results were observed in the other vacuum extrac-
                                                                      8000-i
                                                                                                  VE - 1
                                                     BENZENE


                                                     TOLUENE


                                                     XTLENES
 O
                         50               100
                         DAYS OF EXTRACTION

                            Figure 7
                  Wellhead Vapor Concentrations
150
                                                                VOLATILE ORGANICS MONITORING REMEDIATION    277

-------
tion wells. Table 3 compares the initial vapor concentrations to
the final concentrations observed each wellhead. The relative
decline in extracted vapor concentrations is expected to be propor-
tional to an aggregate soil concentration within the radius of in-
fluence of each vacuum extraction well. Overall, the relative degree
of cleanup achieved by the vacuum extraction process during the
demonstration period, based on initial and final wellhead concen-
trations of total hydrocarbons, ranged from 95.9 to 99.7%.

                           Table 3
                Relative Cleanup Levels Achieved
       During the Vacuum Extraction Demonstration Project

           TOTAL HYDROCARBON CONCENTRAION AT WELLHEAD
Well
          VE-1     VE-2    HE-1-13   HE-l-SO    ME-2-16  HF.-2-58
 Initial:
   (ppซ0   13,900   20,500   16,100    30,700    78,800    27.600
 Ending:
   (ppซ)      570
 (7/24/87)
    52
                      85
                              620
                                      472
Relative Cleanup
Achieved:
           95.91    99.71    99.71    99.71    99.21    98.31
                BEHZENE  CONCENTRATION AT WELLHEAD
Well
          VE-1
Initial:
  (pp.)   4,600
VE-2


 490
Ending:
  (pp.)    0.01     0.16
(7/24/87)
ME-1-13  ME-1-50    ME-2-16   HE-2-58


  730      740        910       360


 0.81     0.25        6.2       ND
Relative Cleanup
Achieved:
          99.999Z  99.971   99.91    99.971

          ND •  leas than 0.01 ppm
                           99.3Z
                                    1001
  Extrapolation of the concentration versus time data is used to
estimate the time-frame for cleanup. Considering the current level
that has been set by the FDER to define excess soil contamina-
tion, 500 ppm of hydrocarbons in headspace of soils, the relative
time-frame to reach this cleanup goal using the Vacuum Extrac-
tion Process can be evaluated. Since the criterion of 500 ppm is
based on a headspace of soil samples using a Flame lonization
Detector (FID) calibrated for methane, the concentration extracted
from  the wellhead may be considered essentially an  aggregate
"headspace" concentration of hydrocarbons (calibrated to gaso-
line) in the soils around the well screen.  Assuming any disparity
between methane standardization and gasoline standardization was
negligible  or at least quantifiable,  the  500 ppm hydrocarbon
response would represent the upper limit of a cleanup goal, where
lower values may be required at certain  sites in order  to protect
ground water resources.
  Using the foregoing discussion as a basis to evaluate the site con-
ditions with regard to the cleanup goal of 500 ppm, the total
hydrocarbons in the extracted vapor in each extraction well were
analyzed using an FID and are shown in Table 3. Considering the
500 ppm headspace concentration as the cleanup goal, the deep
wells  VE-3 and ME-1,  50 are well below the criterion, while
ME-2-58 is slightly under the limit. Similarly, the shallow extrac-
tion wells VE-1 and ME-2 are just above the goal while ME-1-13
is substantially below the cleanup goal. Extrapolation of these data
suggest that the cleanup criterion would  be met on a stable basis
within about six weeks of further vacuum extraction.

278    VOLATILE OROANICS MONITORING REMEDIATION
  An alternative criterion for considering the cleanup goal would
be based on a drinking water standard for indicator parameters
benzene, toluene, ethylbenzene and xylenes. Since benzene has the
lowest MCL in Florida, (1 ug/1) goals for cleanup of the vadose
zone may be considered using the concentration of benzene in the
soil water just above the water table and Henry's Law to calculate
the vapor concentration.
  Based on Henry's Law Constant, the benzene concentration in
air in equilibrium with water having a concentration of 1 ug/1
would be about 0.05 ppm benzene in air. Assuming the extracted
vapors from the wellheads are sufficiently close to equilibrium con-
ditions, an area within the radius of influence of the well may be
considered clean if the concentration of extracted benzene vapors
is less than 0.05 ppm.
  Applying this concept to evaluate a cleanup criterion for wells
extracting benzene near the water table, the data presented in Table
3 would  be used to evaluate the time-frame for cleanup.  This
criterion  probably would be  most  applicable to wells extracting
benzene from near the water table.  Accordingly, ME-2-58  has
surpassed this cleanup criterion, however, VE-2, and ME-1-50
would require further extraction to meet the cleanup goal. The time-
frame for cleanup of the vacuum extraction wells  near the water
table  is estimated to be on the order of 1 month.
  However, if significant hydrocarbons are present above the clay
layer beyond the radius of influence of the extraction wells in the
shallow zone, continued recharge of hydrocarbons to the lower
wells could extend the cleanup time frame. Although considering
dilution, dispersion and adsorption of contaminants migrating from
the shallow zone above the clay layer to the aquifer, a higher
threshold of benzene concentration in extracted vapors from above
the clay layer may be considered to protect groundwater quality
beneath the site.
                                        CONCLUSIONS AND RECOMMENDATIONS
                                          Based on the results of the pilot test, the Vacuum Extraction
                                        Process has been effective in cleaning up the residual hydrocarbons
                                        beneath the site. Testing of subsoils indicated elevated levels of
                                        gasoline hydrocarbons above a clay layer occurring at depths of
                                        15 to 20 ft that are spread at least 50 ft from the source area of
                                        the leaky pipeline. In addition,  high levels of hydrocarbons were
                                        observed near the water table.
                                          Rates of extraction of hydrocarbons from the subsoils using the
                                        vacuum extraction process reached a maximum of 2000 Ib/day
                                        More than 22,000 Ib of gasoline hydrocarbons have been recovered
                                        during the pilot test and subsequent vacuum extraction operations.
                                        As  a result of vacuum extraction operations, hydrocarbon con-
                                        centrations in the groundwater aquifer have been reduced signifi-
                                        cantly.
                                          The vacuum process operated a total of 128 days over a period
                                        of 6 months achieving significant  reductions of  hydrocarbons
                                        within the vadose zone. Extrapolation of the data indicates that
                                        about 4 to 6 weeks of continued vacuum extraction would achieve
                                        the cleanup goals that have been considered for the site.
                                          Recommendations include:

                                        • Sampling of soils within the radius of influence of the vacuum
                                          extraction wells to verify that the cleanup objectives have been
                                          achieved
                                        • Sampling of soils  above the clay  layer and beyond the  radius
                                          of influence of the shallow extraction wells to define the full
                                          lateral extent of the subsurface hydrocarbons beneath the site
                                        • Expanding the  vacuum extraction  system  to clean up  any
                                          hydrocarbons observed in soil tests
                                        • Monitoring groundwater quality to determine the long term
                                          impact the vacuum extraction process has had on improving the
                                          groundwater quality at the site

-------
REFERENCES                                                         Assessment Draft Report," ESE No. 84-521-0200-2140,  1984.
1. Patton, R. H.,  "Florida Department of Environmental Regulation        3. Environmental Science and Engineering, Inc., "Additional Contamina-
  memo Groundwater-City  of Belleview/Union 76  Station, Marion          tion Assessment, Belleview/Union 76 Site," ESE No. 87-207-0100-2150,
  County,"  Volatile Organics Analysis (Gasoline)—File reference T.I5          1987.
     g>     '                                                          4. Malot, J. J., "Preliminary Report, Belleview Subsurface Hydrocarbon
2. Environmental Science and Engineering, Inc., "Belleview Contamination          Cleanup," Terra Vac Inc.. San Juan, Puerto  Rico, 1987.
                                                                  VOLATILE ORGANICS MONITORING REMEDIATION    279

-------
                      Analysis Methods and  Quality  Assurance
                           Documentation  of Certain Volatile
                Organic  Compounds  at Lower Detection  Limits
                                                Susan  Sherman
                                                 CH2M HILL
                                               Denver,  Colorado
                                                 Ward Dickens
                                                  Harold Cole
                                                 CH2M HILL
                                            Montgomery, Alabama
ABSTRACT
  Data gathered during Superfund remedial investigations must
be quality controlled and assured and court defensible if the site
is enforcement sensitive. Data are court defensible if  they are
analyzed according to U.S. EPA methods. However, circumstances
arise when analysis of compounds at lower detection limits than
established by a U.S. EPA method is required to evaluate siie
hazards.
  Recently, nine volatile organic compounds at lower detection
limits were analyzed  for a U.S. EPA Region VI Superfund site.
The compounds are all known carcinogens, and the site is in the
middle of a municipal well field; therefore, the U.S. EPA requested
that these compounds be analyzed at 10'* lifetime excess cancer
risk concentration detection limits. To provide quality data at these
lower detection limits, modifications to U.S. EPA Method  624 were
made to enhance the sensitivity, precision and accuracy of the
method at lower detection limits. Modifications applied to sample
size, matrix spikes, validation of the minimum detection limits,
calibration curves  and daily continuing calibrations.
  The new modified U.S. EPA Method 624 was validated during
a sampling event at the site. The modified U.S. EPA Method 624
data generated 90 to 95% confidence limits, which was acceptable.
 INTRODUCTION
  The U.S. EPA's CLP routinely analyzes many compounds by
 U.S. EPA methods; however, the standard detection limits for these
 methods may be above those required for a thorough site evalua-
 tion.  Consequently,  other methods  must  be  used  following
 protocols that maximize their utility and  defensibility.

 BACKGROUND
  The San Jose 6 Superfund site was placed on the NPL in 1981,
 and U.S. EPA studies at the site were initiated in 1984 in accordance
 with CERCLA. The contamination of groundwater produced by
 Well San  Jose 6 caused it to be taken out of service in 1980, and
 an immediate remedial measure (IRM) was initiated to provide a
 replacement well which is  under construction.
  The site is located in the South Valley area of Albuquerque,
 Bernalillo  County,  New  Mexico. The site, located within  an
 industrial and residential area, encompasses approximately 2 sq
 mi. It has been designated as the state's highest priority Superlund
 site because hazardous substances are present in the groundwater
 in the City's  San  Jose well field.
  The remedial investigation/feasibility study is focused on existing
 and historic contamination of Well San Jose 6. The overall objec-
 tives of remedial action are to protect the public health and welfare
and the environment  by protecting  the undivided  Santa Fe
Formation aquifer in  the  vicinity  of Well  San Jose 6 from
contamination.

CONTAMINANT EVALUATION
  Groundwater contaminants historically encountered at the San
Jose 6 Superfund site include a broad  suite of identified volatile
organic  compounds and aromatic petroleum  products. The
following list presents a summary of water quality data from Well
San Jose 6 and  four wells on PRP properties. The following list
presents only the five highest contaminants reported in the most
contaminated well in each area:
• Well San Jose 6 (deep aquifer):

      Contaminant
1,1,1 -Trichloroethane
1,1-Dichloroethane
1,1-Dichloroethene
Trichloroethene
Tetrachlorethene

• PRP 1

      Contaminant
4-Methyl-2-Pentanonc
1,2-Trans-Dichloroethane
1,1-Dichloroethane
Methylene Chloride
Tetrachloroethene

• PRP 2

      Contaminant
N-Methylnaphthalene
Xylene
Phenanthrene
N-Nitrosodiphenylamine
Benzene

• PRP 3

      Contaminant
1,1-Dichloroethane
1,1-Dichloroethene
Trichloroethene
4-Methyl-2-Pentanone
Tetrachloroethene
Concentration 0
-------
• PRP 4

       Contaminant
Pentachlorophenol
Toluene
Tetrachloroethane
1,2-Dichloroethane
Concentration (jig/1)
       179
        34
        12
         5.8
  In addition, levels of contamination near or below the detec-
tion limit were reported throughout the area. Analyses had been
performed  in the numerous studies  by  different parties using
different methods of analysis and different levels of detection.
Initial U.S. EPA studies utilized the U.S. EPA CLP for sample
analysis. This program uses standardized methodologies, quality
assurance/quality control and reporting for consistency and defen-
sibility.

PROBLEM DEFINITION
  During the planning stage for the most recent U.S. EPA field
studies at the San Jose 6 Superfund site, the potential risk because
of exposure to the contaminants at the site was evaluated to provide
guidelines for remedial actions. Remediation guidelines often used
for carcinogen and  suspected carcinogen contamination are the
10~6 cancer risk concentrations. It was noted that CLP routine
analytical services (RAS) detection limits significantly exceed the
10 "6 cancer risk concentration for some contaminants of concern
at the San Jose 6 site.  An attempt  was made via the Special
Analytical  Services  (SAS) of the CLP to utilize more sensitive
analytical techniques. A suitable U.S. EPA analytical protocol
(Proposed Method 524.1) was identified by William Langley of
the U.S. EPA's Houston laboratory. Bids were solicited from par-
ticipant laboratories in the CLP  to analyze selected samples
collected from the site using this protocol. No laboratories within
the CLP were interested in performing the SAS analyses for the
San Jose 6 site as specified in the invitation to bid. One labora-
tory, California Analytical Laboratories (CAL), responded that
it could  perform such analyses but only if allowed to use a modified
version  of U.S. EPA  Method 624.
  After  further consideration,  alternatives  for  resolving this
problem were developed, as listed below:

• Perform analyses utilizing U.S. EPA Method 624. This  alter-
  native would not yield analytical results that would be sensitive
  in the "decision range" (10~6 cancer risk concentration). This
  particular deficiency in the 624 Method had already been brought
  to the attention of the U.S. EPA by one of the PRP's at the site.
  On the other hand, analyses by this  method could be performed
  by the CLP program readily,  without delaying the field effort.
  Analyses would be  as per an established U.S. EPA method.
• Perform analyses utilizing modified  U.S. EPA Method 624. This
  method would provide the means to achieve the desired sensi-
  tivities. The only departure from U.S. EPA Method 624 would
  be that a lO-milliliter (ml) sample,  rather than a 5-ml sample,
  would be used. With this modification, the protocol would be
  identical to proposed Method 524.1 except that purge sample
  volumes and internal standards differ. Three compounds would
  be used as internal standards, as opposed to one internal standard
  for proposed Method 524.1, and the single internal standard used
  in proposed Method 524.1 would not be among the three inter-
  nal standards used in modified Method 624. Only minor delays
  would be experienced in the field effort since some time would
  be required for the U.S. EPA to incorporate suitable quality con-
  trol measures into the revised bid. Analyses would not be as per
  an established U.S. EPA method, which would make  court
  defensibility of results more difficult.
• Procure the services of an analytical lab outside the CLP and
  use proposed Method 524.1. If a laboratory could be found that
  would use proposed Method  524.1, desired sensitivities would
  be achieved and approved U.S. EPA methods would be used for
  the analyses. Some delays in field effort would be experienced
  while a qualified laboratory is located. Questions on laboratory
  qualifications may arise since such a laboratory would not have
  been subject to thorough CLP screening procedures.
• Bring a laboratory into the CLP that  could use Method 524.1.
  as the third alternative above. The project schedule would be
  adversely affected in that as many as 6 months may be required
  to bring a laboratory into the CLP.
• Use approved, appropriately sensitive U.S. EPA Gas Chromato-
  graph (GC) Methods 601 and 602 for analyses. It should be
  possible to achieve desired sensitivities,  and established U.S. EPA
  analytical methods would be used. Project schedules should not
  be delayed too much since  many laboratories have GC capa-
  bility for such analyses. The non-confirmatory nature of GC
  analysis (versus GC/mass spectrometer analyses used in Methods
  524.1 and 624) would likely cast some doubt upon the interpre-
  tation of analytical results. This alternative was bid  through the
  SAS component of the CLP; however, no one responded to the
  request for bids.

SELECTED METHOD
  To achieve the desired detection limits for volatile  compounds
of concern, the modified U.S. EPA Method 624 was the second
alternative selected of those discussed above. These analyses will
be performed using the SAS component of the CLP so that all
RAS CLP contractual obligations apply  to this method except for
the  modifications discussed below.
  U.S. EPA Method 624, dated July 1982, is  an RAS CLP method.
The purposes of these modifications to this method are to enhance
the  sensitivity,  precision and  accuracy  of  the method at lower
detection limits. The modifications include the following:

• Sample size:  Increase from 5  ml in U.S. EPA Method 624 to
  25 ml. This change increases  the sensitivity of the method of
  factor by five.
• Matrix spikes: Use the U.S. EPA contract  specified spiking com-
  pounds. Decrease the matrix spike and matrix spike duplicate
  concentrations from 20 /tg/1 per parameter as outlined in U.S.
  EPA Method 624 to 1 ^g/1 per parameter as the bottom point.
  This change will demonstrate the precision and accuracy of the
  method at lower concentrations by better estimating the detec-
  tion  limits. The field team shall provide  additional samples at
  a frequency of one well in 10 to accommodate the  laboratory.

• Validation of the minimum detection  limit (MDL):  To validate
  the sensitivity at lower concentration  levels, three standards at
  a concentration of 1.0 /tg/1 shall be analyzed and the standard
  deviation determined. Multiply the standard deviation by three to
  obtain an estimate of the instrument detection limit.

• Calibration curves: Establish the linearity of the instrument in
  the range of 1.0 to 100 jtg/1. At least five calibration points must
  be acquired with the lowest point being 1.0 /tg/1. Concentrations
  of 1.0, 10, 20, 50 and 100 /ig/1 shall be  analyzed  to establish
  linearity in this range.


                            Table 1
    Compounds to be Measured Under San Jose 6 Site SAS Request
                             Volatile Fraction/Compound

                             Benzene
                             Carbon tetrachloride
                             Chloroform
                             1,2-dichloroethane
                             1,1,2,2-tetrachloroethane
                             1,1,2-trichloroethane
                             Trichlorethylene
                             Tetrachloroethylene
                             Vinyl chloride
                             DE/SVLY7/038


                        VOLATILE ORGANICS MONITORING REMEDIATION    281

RAS
Detection
Limits
(microqrams
per liter)
5
5
5
5
5
5
5
5
10
10~6 Life-
time Excess
Cancer Risk
Concentrations
(micrograms
per liter)
0.67
0.42
0.19
0.94
0.17
0.6
2.8
0.86
1.0


Detection
Limits
fmicrograms
per liter)
0.21
0.13
0.20
0.35
0.2
0.5
0.18
0.5
0.5

-------
• Daily continuing calibrations: Use the 10 jig/1 standards for the
  daily continuing calibrations. The response factors of the U.S.
  EPA specified continuing calibration compounds (CCC) must
  not deviate more than 30%  from the mean of the calibration
  curve.

  Table 1 presents the detection limit targets for the compounds
of concern.
  In addition, samples will be analyzed using  RAS methods to
maintain court defensibility of the data base.

METHOD VERIFICATION
  This method was evaluated during a preliminary groundwater
sampling round in May 1986. This evaluation consisted of analysis
of seven replicates from two wells using RAS and  modified Method
624 techniques.
  Five compounds (trichloroethene, benzene, tetrachloroethene,
1,2-dichloroethane and chloroform) were detected by the modi-
fied method. The RAS method detected only trichloroethene in
samples from one well. Analytical results are shown in Tables 2
through 6.
  Both methods were reasonably consistent in their detection and
values reported. Confidence limits for the modified method results
ranged from 90^0 ฑ0.08 /ig/1 to 98% ฑ2.81 /ig/1 for different
compounds as shown in Table 7.
                           Table 2
                  Trichloroelhene Analyses (/ig/l)
                         1986 Sampling
    Sjr.ple
                                       Well CVD-1
Original
Repl icate
Repl icate
Replicate
Replicate
Replicate
Repl icate
Replicate
Mean Value

1
2
3
4
5
6
7

0.
0.
0.
0.
0.
.2
.2
•ป
2
2
ND
0.
2
ND
0.
15
ND
ND
Nn
ND
ND
ND
ND
ND

7.
8 .
9.
10.
10.
11.
8.
7.
9.
1
6
4
2
3
7
0
6
11
9
5
6
5
2
10
9
a
6.75
   Modified Method 624.

   Routine Analytical  Services, Method 624.

  Sote:   ND • not detected.


  DE/SVLY7/030.1
Table 3
Benzene Analyses (jig/I)
1986 Sampling
Sample
Or iq ina 1
Replicate
Replicate
Replicate
Replicate
Replicate
Replicate
Replicate
Mean Value


1
2
3
4
5
6
7

Wei 1
0.
0,
0,
0.
0.
0.
0.
0.
0.
A-l"
, 3
. 3
,3
.2
,2
2
2
2
24
RASb
NIi
ND
ND
ND
ND
ND
ND
ND

well
0
0
0
0
0
0
0
0
0
CVD-1 *
.2
.2
.2
.2
.2
.2
.1
.1
.18
                                                       RAS

                                                        ND
                                                        ND
                                                        ND
                                                        ND
                                                        ND
                                                        ND
                                                        ND
                                                        ND
  Modified Method 624.
  Routine Analytical Services,  Method 624.

 Note:  ND -  not detected.


 DE/SVLY7/030.2


282    VOLATILE OROANICS MONITORING REMEDIATION
                                                                      Sample
                                                                                             Table 4
                                                                                  Tetrachloroethene Analyses
                                                                                          1986 Sampling

                                                                                    Well A-l'
Or iqi na 1
Rep 1 icate
Repl icate
Repl icate
Repl icate
Replicate
Replicate
Repl ica te
Mean Value

1
2
3
4
5
6
7

0
0
0
0
0
0
0
0
0
.6
.6
.6
.5
.5
.4
.4
.4
.50
U\i
ND
ND
ND
ND
ND
ND
ND

Hell CVD-1

    0.5
    0.6
    0.6
    0.6
    0.6
    0.7
    0.6
    0.5
    0.59
                                                                    Modified Method  624.
                                                                   bRo,,t|ne Analytical  Servicei, Method 624.

                                                                   Note:  ND • not detected.
                                                                   DE/SVLY7/030.3
                                                                                             Table 5
                                                                                  1,2-Dkhloroelhane Analyses (/if/I)
                                                                                          1986 Sampling
                                                                                   Well A-l0
Original
Replicate 1
Replicate 2
Replicate 3
Replicate 4
Replicate 5
Replicate 6
Replicate 7
Mean Value
'Modified Method
Routine Analytic
25. 1
25.2
24.6
34.0
32.1
35.0
32.9
33.0
30.26
624
ral Services,
ND
ND
ND
ND
ND
ND
ND
ND

Method *24.
ND
ND
ND
ND
ND
ND
NL
ND


                                                                                                          Well CVD-la
                                                                   DE/SVLY7/030.4
Table 6


Sample
Or iq inal
Replicate 1
Replicate 2
Replicate 3
Replicate 4
Replicate 5
Replicate 6
Replicate 7
Mean Value
Chloroform
1986
Well A-l'
NT
KC
ND
ND
ซ[
ND
[ID
ND

Analyses
Sampling
RAS5
VD
NT
SD
ND
ND
ND
ND
ND

(Kg 1)

Well CVD-1 "
ND
0.6
0.6
0.6
0.6
0.6
ND
ND
0.38
                                                                    Modified Method 624.
                                                                    Routine Analytical  Services, Method 624.

                                                                   Not*:  ND • not detected.


                                                                   DE/SVLY7/030.5


                                                                                             Table 7
                                                                                  Summary of Method Evaluation
                                                                                       Modified Method 614
                                                                                        May 1986 Sampling
                                                                                                                          NO
                                                                                                                          ND
                                                                                                                          NO
                                                                                                                          ND
                                                                                                                          NO
                                                                                                                          HO
                                                                                                                          ND
                                                                                                                         ND
                                                                                                                         NO
                                                                                                                         ND
                                                                                                                         no
                                                                                                                         ND
                                                                                                                         ND
                                                                                                                         ND
                                                                                                                         ND
                                                                                                                        RAS"

                                                                                                                         ND
                                                                                                                         ND
                                                                                                                         ND
                                                                                                                         ND
                                                                                                                         ND

                                                                                                                         ND
                                                                                                                         sr
                                                                                    Compound
                                                                              Trichloroethene
                                                                              Benzene
                                                                              Tetrachloroethene
                                                                              1,2-Dichloroethane
                                                                              Chloroform
                                                                                                          Confidence Limits
98ซ
90%
95ป
95%
95ป
2.
0.
0.
6.
0.
81
08
2
1
35
ppb
ppb
ppb
ppb
ppb
                                                                    DE/SVLY7/030.6

-------
CONCLUSION
  For the San Jose 6 Superfund  site, analytical requirements
included modifications to U.S. EPA Method 624. The purposes
of the modifications to this method were to enhance the sensitivity,
precision and accuracy of the method at lower detection limits.
Modifications applied to sample size, matrix spikes, validation of
the minimum detection limits, calibration curves  and daily con-
tinuing calibrations.
  The new modified U.S. EPA Method 624 was validated during
a sampling event at the site. Ten wells were sampled; seven repli-
cates plus an original sample were collected from two wells; volatiles
were analyzed by RAS and SAS methods. Statistical analysis was
done on the replicate data to determine laboratory quality esti-
mation. The modified U.S. EPA Method 624 data generated 90
to 95% confidence limits, which were of acceptable quality. The
RAS volatile samples will be analyzed when SAS volatile samples
are collected at this  site  to provide the court defensible data
requirement for enforcement sensitive Superfund sites.
                                                             VOLATILE ORGANICS MONITORING REMEDIATION    283

-------
         Toxic  Air  Quality  Investigation at a  Hazardous  Waste  Site
                                                George D.  Marquardt
                                                      CH2M HILL
                                                Milwaukee, Wisconsin
INTRODUCTION
  Ambient air quality monitoring at NPL sites for chemical con-
taminants is an important part of each site investigation. This paper
will present ambient air quality and soil gas data obtained during
an investigation of a hazardous waste site currently being operated
as a municipal landfill. The primary objective of this program was
to identify and quantify chemical contaminants being emitted by
the site via the air pathway.  Because this is a CERCLA enforce-
ment site,  no  reference will be made to its identity or location.

Site History
  The property was acquired from the U.S. government in 1964
and in began  operation  in 1965 as a sanitary landfill. Municipal
and industrial wastes were placed in excavated pits. The industrial
wastes were liquids and solids that are now classified as hazardous
wastes. The waste pits, ranging from 15 to 40 ft deep with varying
widths and lengths, were filled one-half to three-quarters full with
liquid waste and then filled with refuse to absorb the liquids. Refuse
was placed to form a cap rising several feet above the original surface
to allow for subsidence because of surface traffic over the waste
pits. This  method of disposal was used from 1965 to 1980.
  Records were not kept of  the liquid  wastes disposed of during
the city's management of this site. No measures were taken to pre-
vent leachate or waste seepage from the pits because there was no
evidence that  such measures were needed.
  Attempts to account for the wastes disposed of at the site indi-
cated the following general categories were deposited in the pits:
acid and alkaline sludges; caustics and solids with pH's below 2
and near 14; brines including plating wastes, and other  water-based
sludges; organics, both natural and synthetic, including petroleum-
based oils, grease, chlorinated solvents  and sludges, water-soluble
oils, municipal sewage sludge,  radioactive wastes  and pesticide
wastes.
  Low-level radioactive wastes from hospitals and small laboratories
were disposed of at the site from 1968 to 1977. From 1977 to 1980,
radioactive wastes were  disposed of only in specified waste pits.
  Disposal of domestic sewage sludge took place in the northern
quarter  of the site from 1970 to 1980.  An oil company disposed
of oil sludges on a 20-acre area  from  1975 to 1980. Vehicle tires
form large mounds that cover several  areas.
  An independent subcontractor assumed control over the waste
disposal operations in August 1980. The landfill operations at the
site now dispose of only municipal refuse; however, asbestos dis-
posal began in late 1983 in trenches on the east side of the landfill.
  In 1977, the USGS conducted a study of the groundwater quality
near the site and concluded  that the disposal activities at the site
were polluting shallow alluvial and upper bedrock groundwater in
the area.
  The groundwater at the site is contaminated with varying con-
centrations of organics.  Contamination is reported in areas adja-
cent to the intermittent stream draining pan of the site and adjacent
to waste disposal pits located at the site.
  In 1984, this site was put on the NPL Priorities List for cleanup
under the Superfund Act because contamination from the site was
suspected to be entering groundwater resources. CH2M HILL was
retained by the U.S.  EPA  to do a RI/FS. The purpose of the RI
was to determine the nature and extent of contamination, assess
the potential for migration of contaminants from the site and assess
the possible public health and environmental effects from the con-
tamination. The FS developed and assessed remedial action alter-
natives to control or eliminate the migration of contaminants from
the site.

Air Quality Sampling
  The original air quality sampling program included two compo-
nents: a short-term program to  monitor ambient air quality con-
centrations at off-site locations  upwind and downwind of the site
during drilling of waste pit well points and a long-term program
to continuously monitor air quality at the same locations for 1 year.
  After review, the U.S. EPA's  Air Quality group revised the air
quality  sampling program in June  1985. The revised program
included the following:

Short-Term Program

• Upwind and downwind sampling to evaluate potential air quality
  impacts associated with the landfill
• Sample at locations close  to (nominally 50  ft) and downwind
  of the site investigation activities involving waste pit well installa-
  tions; these data  are to be used  to assess air quality impacts
  associated with the site investigation activities
• Sample at locations close to and downwind of the waste pit gas
  well during a controlled release in which the waste pit gas well
  would be permitted to naturally vent itself; these data are to be
  used by the U.S. EPA's Air Quality group to predict ambient air
  impacts associated with waste pits at the site

Long-Term Program
• A long-term program to be  designed  and implemented  if
  warranted by results of the short-term program.

  Wind direction and therefore on-site  sampling locations were
determined at the meteorological station near the on-site command
post.
284    VOLATILE OROANICS MONITORING REMEDIATION

-------
  Air quality sampling was conducted at the site during a 12-day
period  in November and early December 1985.  Samples were
collected at the upwind, downwind and on-site locations for vola-
tiles, semivolatiles, metals, pesticides, PCBs and total suspended
particulates (TSP). During  the  sampling  period, the following
samples were collected:

• Fifty-two Tenax samples for highly volatile organic compounds
• Fifty-two carbon molecular sieve (CMS) samples for volatile
  organic compounds (VOCs)
• Twenty high-volume air filter samples for TSP and metals
• Twenty-six polyurethane foam (PUF) samples for semivolatiles
  and pesticides/PCBs

Sampling Equipment
  At the upwind and downwind air quality monitoring locations,
air samples were collected by trapping them on carbon molecular
sieve (CMS), polyurethane foam (PUF), Tenax and glass fiber filters.
At the on-site and controlled release stations, samples were collected
on CMS and Tenax material only.
  Sampling equipment consisted primarily of the following:

• Gillian HFS battery-operated personal sampling pumps were used
  to collect all Tenax and CMS samples at the upwind, downwind,
  on-site and controlled release locations.
• Sierre Accu-Vol high-volume TSP samplers were used at the up-
  wind and downwind locations to collect TSP and  airborne
  metal samples.
• General Metal Works (GMW) Model  PS-1 high  volume air
  samplers were used at both the upwind and downwind locations
  to collect PUF samples.
• A Climatronics Wind Mark III Meterological Station was used
  at the on-site command station to collect and record meteoro-
  logical data.
• Tenax and CMS cartridges, which adhered to the specifications
  in the Special Analytical Services (SAS) Regional Request forms
  and  to Methods TO-1 and TO-2, respectively, were used  at all
  locations.

  Procedures for operating,  maintaining and  calibrating  the
sampling  equipment were  according  to the  manufacturer's
guidelines.
  The Gillian HFS sampling pumps were calibrated and leak tested
using the Gillian calibrator each day of use The flows were recorded
before and after each sample period with the appropriate cartridge
(either Tenax or CMS) in line.
  The GMW PS-1 PUF samplers were calibrated prior to sampling
using a fixed orifice and manometer with a blank cartridge in line.
Calibration curves of the magnehelic gauge versus  manometer
readings  (in. H20)  and actual flow versus  magnehelic gauge
readings were prepared and used to record actual daily flows.
  The Sierre high-volume TSP samplers were calibrated using a BGI
variable orifice and manometer. The actual flows were corrected
to standard temperature and pressure.

Sampling Procedures
  All samples were handled,  shipped and documented as described
in the QAPP. Each PUF sample submitted to the CLP laboratory
for analysis was wrapped in the original foil wrap as required. PUF
glass cartridges were used only once.
  All TSP filters were folded in half and  then in half again and
placed  in Ziploc bags. The filters were held and shipped once a
week to the CLP laboratories, whereas the Tenax, CMS and PUF
samples were shipped daily.
  Sampling procedures consisted primarily of collecting Tenax,
CMS, PUF and  high-volume filter samples at the upwind and down-
wind locations at specific flowrates for each day that sampling took
place. The samplers were run from 8 to 12 hr, depending on the
duration of daily investigation activities. Flowrates varied from 100
to 500  ml/min  for the samples.
  TSP  and airborne heavy metal samples at the upwind and down-
wind locations were collected each day during this period by drawing
a known volume of air through an 8- by 10-in. glass fiber filter.
  Pesticide and PCB samples were collected at the upwind and
downwind locations each day during the air quality sampling period
by drawing a known volume of air through a 6-in.  diameter glass
fiber filter and a PUF plug.  Air flow through the system was
monitored by a venturi/magnehelic assembly.

Analytical Procedures
  Analytical procedures used for CMS, Tenax, PUF and TSP
samples were as follows:

CMS Samples
  Fifty-two CMS samples were analyzed for all highly volatile VOCs
on the HSL using U.S. EPA Method TO-2, "Method for the Deter-
mination of Volatile Organic Compounds in Ambient Air by Car-
bon Molecular Sieve Adsorption and Gas Chromatography/Mass
Spectrometry (GC/MS)"1.

Tenax Samples
  Fifty-two Tenax samples were analyzed for all VOCs on the HSL
using U.S. EPA Method TO-1,  "Method for the Determination of
Volatile Organic Compounds in Ambient Air by Tenax Adsorp-
tion and Gas Chromatography/Mass  Spectrometry (GC/MS)"1

PUF Samples
  Twenty-six PUF samples collected at the upwind and downwind
locations during the air quality investigation were analyzed for all
HSL semivolatiles (base/neutrals and acids) and pesticides/PCBs.
These samples were solvent extracted according to U.S. EPA Method
625. The sample extract was divided into two halves. One half was
analyzed for all HSL semivolatiles using GC/MS Method 625. The
other half of the extract was analyzed for pesticides/ PCBs  by sol-
vent exchange with hexane and GC/ECD analysis according to U.S.
EPA Method TO-4, "Method for the  Determination of Organo-
chlorine  Pesticides and Polychlorinated  Biphenyls  in Ambient
Air"1.

TSP and Metal Samples
  Twenty high-volume TSP filter samples were collected at the up-
wind and downwind locations during this period.
  The  laboratory preconditioned the ambient filters in a con-
ditioning chamber for 24 hr at 70 ฐF (plus or minus 5ฐ) and 50%
humidity (plus  or minus 5%). The filters then were preweighed and
shipped to the  site. The preconditioning and weighing procedures
adhered to U.S. EPA guidelines for TSP  monitoring2.
  The exposed filters were sent back to the same laboratory that
performed the original weighing  for final  conditioning  and
weighing. The final conditioning and weighing procedures were the
same as the original procedures. These filters then were sent to the
CLP laboratory for analysis of all HSL metals, as indicated in the
guidelines.
  Metals (including lead) and cyanide analysis on the high-volume
filters used the routine analytical services detection limits and proce-
dures as specified in the U.S. EPA guidelines. Analysis was done
on a 1.5 in. strip obtained from each filter, as per the EPA Refer-
ence Method for lead3.                                   u

Air Sampling Results
  There currently are no U.S. EPA or state ambient air quality
standards for HSL volatile organic compounds, semivolatiles, metals
(except lead) or pesticides and PCBs. Consequently, discussion of
ambient air quality concentrations of these compounds centers
around upwind (background) concentrations versus downwind con-
centrations from the landfill.
  The difference between these two demonstrate potential landfill
contributions of these compounds to the ambient air quality.

Tenax Sampling Results
  Sample volumes for each sample were calculated and corrected
to standard temperature and  pressure (STP)  using the  on-site
                                                              VOLATILE ORGANICS MONITORING REMEDIATION     285

-------
meterological  data. All blank  Tenax samples  were reported in
ng/tube because air was not allowed to pass through the tubes. The
legend for Tables 2 through 14 is  shown in Table 1.
                             Table  1
                   Legend for Tables 2 through 14
   For the purposes of this data  review document the following
   code letters and associated definitions are provided.
                            Table 2
Volatile Organic Compound Concentrations for Tenax Field Blank Sampfe
                   IMtT-lo  imf-M   >ซปT-il   HOซ-ซ!  tm-t-M  UBU
                   nnj> tu  FIELD iu  noA au r\xu> ft*  rin/>
                          lt-lป-ซ   IJ-
        U -  The  material was analyzed for, but was not
             detected.  The  associated numerical value ii the
             estimated  sample quantitatlon limit.

        J    The  associated  numerical value is an estimated
             quantity because the amount detected is below the
             required limits or because quality control
             criteria were not met.

        UB   Estimated  sample quantitatlon limit increased.
             Amount  found in sample reported.  Compound
             detected at less than 5 times the amount in blank
             (less than 10 times for methylene chloride,
             acetone, toluene and phthalatea).

        UJ   Detection  limit is estimated because quality
             control criteria were not met.

        JB   The  value  is an estimated amount detected below
             required limits and also detected in thi- blank.
        n    Compound was detected in the blank.  Quantity
             reported is qreatvi  than 5 times  the amount  found
             in the blank (greater  than 10  times  for methylene
             chloride, acetone,  toluene,  and  phthalates).

        R -   Quality Control indicates  that data are not  usable
             (compound may or may not  be  present) .  Resampling
             and  rcanalysia are  necessary for  verification.

        Z    No analytical  result.

        N    Presumptive evidence of  presence  of material
             (tentative  ident if icat ionl .

   Concentrations were reported for 35 compounds in the volatile
 organics category. Five field blanks were taken and showed varying
 amounts of methylene chloride, acetone,  carbon disulfide and
 toluene, as listed in  Table 2.  Although the field and batch blanks
 showed various  amounts of contaminants, concentrations  were
 generally three to five times less than amounts found in the samples
 and thus considered  acceptable. Constituents  for 10  samples
 collected at the upwind station are reported in Table 3 in average
 concentrations along with several tentatively identified compounds
 (TICs). The average amounts of 4.4 ppb carbon disulfide, 2.5 ppb
 toluene and 1 ppb methylene chloride identified are similar to quan-
 tities also found in Tenax tube blanks.
 286    VOLATILE ORGANICS MONITORING  REMEDIATION
—
Ckl.ro.. K—
N.t*rin otMio*
^..*
CซrซM DiMin*
• 1, |-O1<*I •**•!ป•ป*
,,„.,.. -.,..,—,—
ป!.„,.,.
'.'••'<•'— —

1,1, l-Ti 1 -Mir**' iMfta
ปinri **.t.ซซ
•rwotf 1 r tit •**•• UMM
"—•'•- 	
oi.Miป<__a_
I.l.l-Mซlซ.tt—
•-.-.
...-l.l-MOI.nOT-.
1-on^.a,. n.i •ป.,
.,_,.~
,-_—.
.*,*,,-*.ซ—.
!• t f •ซ•> 1 *fM tkMM
T.tMM
ป,.ซ*ซ._
!ซ„ ^ 	
.„ซ*.
>ซ.l U.™.
mi.i
10 0
10 0
10 t
• J
10 •
10 •
w t
to *
10 •


to t
10 •
W 0
to t
to r
10 IT
10 •
10 *
10 t
!• t
10 V
10 B
• J
to v
10 W
10 W
10 B
ปป
10 • 10 B U •
10 I 10 i to B
40 IK 100 B
10 B 1*0 • 10 B
M H WOO >
10 0 10 * WO
10 f 10 B MB
10 t 10 B 10 B
W t IO B 10 *


1O B 10 It U ซ
tO 0 10 9 10 •
tO G tO B 10 0
10 • 10 B 10 B
to o to B to e
I* I* B 10 b
IO • 10 B IO 0
10 0 10 0 IO V
10 V U 0 10 0
10 ซ to o to a
ซO 10 0 10 V
11 10 b 10 V
IM 100 MO
10 0 10 U 10 0
10 u 10 u 10 r
10 B 10 V 10 B
O 10 IF 10 0
*%•• U* W41
10 f
Ifi t
\y. •
10 •
l*j
10 •
10 0
10 f
10 1


U B
10 f
10 0
10 •
10 ซ
10 0
10 B
10 *
to •
10 I
10 0
10 0
M
10 0
10 0
10 .
10 0
^ n
1C
10
IM
M
IM;
10
to
U
to


10
10
10
10
10
16
11
U
10
U
IB
U
11
1H
10
10
1C
I'
IT*:
   rwunvtu unririo a**u
                                   10 ",J
                                   too P.J
                                   n B,J
                                   *OO ซ,J
  The downwind samples listed in Table 4 show average concen-
trations of 16 ppb carbon disulfide, 12 ppb toluene, 3 ppb acetone
and 2 ppb methylene chloride In addition, traces of benzene, tri-
chloroethene and 1,1,1  trichloroethane were detected. These con-
centrations generally are .1 to 4 times higher than the upwind samples
and 5 to 10 times  higher than the blanks.

-------
                                                               Table 3
                                            Volatile Organic Compound Concentrations for
                                                     Tenax Upwind Air Samples
SAMPLE NUMBER
SAMPLE TYPE
                              1905T-01  1905T-03  1905T-11   1905T-13  1905T-17  1905T-26  1905T-29  190ST-37   190ST-44  1905T-49  AVERAGE

                                                                                                                              UPWIND

                              UPWIND    UPWIND    UPWIND    UPWIND    UPWIND    UPWIND    UPWIND    UPWIND    UPWIND    UPWIND
SAMPLE DATE
                              11-19-85  11-20-85  11-12-85  11-22-85  11-23-85  12-03-85  12-04-85  12-05-85   12-06-85  12-07-85
AMBIENT AIR CONSTITUENTS
                              ppb
                                       ppb
                                                 ppb
                                                           ppb
                                                                    ppb
ppb
ppb
                                                                                                 ppb
ppb
ppb
                                                                                                                              ppb
Chloromethane

Bromomethane

Vinyl Chloride

Chloroethane

Methylene Chloride

Acetone

Carbon Dlsulfide

1,1-Dichloroethene

1,1-Dichloroethane

trans-1,2-Dlchloroethene

Chloroform

1,2-Dichloroethane

2-Butanone

1,1,1-Trichloroethane


Carbon Tetrachlorlde

Vinyl Acetate

Bromodichloromethane

1,2-Dichloropropane

trans-l,3-Dlchloropropene

Irichloroethene

Chlorodlbromomethane

1,1,2-Trichloroethane

Benzene

cls-l,3-Dlchloropropene

2-Chloroethyl  Vinyl Ether

Broraoform

2-Hexanone

4-Methyl-2-pentanone

Tetrachloroethene

1,1,2,2-Tetrachloroethane

Toluene

Chlorobenzene

Ethyl Benzene

Styrene

Total Xylenes

          TOIAL
0.083
0.044
0.068
0.065
1.043
0.073
3.158
0.044
0.043
0.044
0.035
0.043
0.059
0.443
0.088
0.049
0.026
0.037
0.038
0.803
0.020
0.032
0.054
0.038
0.039
0.017
0.042
0.042
0.025
0.025
1.832
0.037
0.095
0.154
0.064
8.80012
U
U
U
U

U

U
U
U
U
U
U


U
U
U
U

U
U
U
U
U
U
U
U
U
U
B
U

B


0.267 U
0.143 U
0.218 U
0.211 U
1.921
0.234 U
1.785
0.140 U
0.137 U
0.140 U
0.114 U
0.137 U
4.337
0.520
0.097
0.158 U
0.083 U
0.120 U
0.123 U
6.621
0.065 U
0.102 U
0.418
0.123 U
0.128 U
0.054 U
0.136 U
0.190
0.041
0.081 U
1.328
0.121 U
0.128
0.287
0.128 U
20.8344
0.027 U
0.041 U
0.040 U
0.030 U
0.417
0.034
0.027
0.093 U
0.027
0.022 U
0.026 U
0.036 U
2.679
0.714
0.030 U
0.016 U
0.023 U
0.023 U
0.020 U
0.196
0.019 U
0.033 U
1.055
0.010 U
0.010 U
0.026 U
0.026 U
0.016 U
0.014
0.028 U
1.538 U
0.024 U
0.025 U
0.024 U
0.024 U
7.39137
0.123 U
0.066 U
0.100 U
0.097 U
3.102
0.108 U
3.214
0.065 U
0.063 U
0.065 U
0.053 U
0.063 U
2.089
1.223
0.041 U
0.073 U
0.038 U
0.056 U
0.057 U
0.096
0.030 U
0.047 U
0.458
0.057 U
0.058 U
0.025 U
0.063 U
0.063 U
0.038 U
0.037 U
1.090
0.056 U
0.059 U
0.060 U
0.059 U
12.9893
0.080 U
0.043 U
0.065 U
0.063 U
0.034
0.070
0.054 U
0.042 U
0.041
0.042 U
0.034 U
0.041 U
0.057 U
0.031 U
0.027 U
0.047 U
0.025 U
0.036 U
0.037 U
0.031 U
0.020 U
0.031 U
0.052
0.037 U
0.038 U
0.016 U
0.041 U
0.041 U
0.025 U
0.024 U
0.044
0.036 U
0.038 U
0.039 U
0.038 U
1.41932
0.066 U
0.035 U
0.054 U
0.052 U
0.039 U
0.104
0.044 U
0.035 U
0.034 U
0.035 U
0.028 U
0.034 U
0.046 U
0.025 U
0.022 U
0.039 U
0.020 U
0.030 U
0.030 U
0.026 U
0.016 U
0.025 U
0.043 U
0.030 U
0.031 U
0.013 U
0.033 U
0.033 U
0.020 U
0.020 U
0.055 N
0.030 U
0.032 U
0.032 U
0.032 U
1.24253
0.064 U
0.034 U
0.052 U
0.051 U
0.223
0.899 N,
0.291
0.034 U
0.033 N,
0.034 U
0.027 U
0.033 U
0.045 U
0.071
0.014 U
0.038 U
0.020 U
0.029 U
0.029 U
0.070
0.016 U
0.024 U
0.092
0.029 U
0.031 U
0.013 U
0.033 U
0.033 U
0.020 J
0.019 U
0.039
0.029 U
0.031 U
0.031 U
0.031 U
2.56051
0.039 J
0.035 U
0.054 U
0.052 U
0.868 B
0.058 U
1.937
0.035 U
0.034 U
0.035 U
0.028 U
0.044
0.191
0.181
0.022 U
0.039 U
0.020 U
0.030 U
0.030 U
0.036
0.016 U
0.025 U
0.378
0.030 U
0.031 U
0.013 U
0.033 U
0.033 U
0.030
0.020 U
4.002
0.030 U
0.025 J
0.032 U
0.082
8.54884
0.077 U
0.042.U
0.063 U
0.104
1.487 B
0.068 U
25.918 >
0.020 J
0.048
0.041 U
0.033 U
0.044
0.055 U
0.592
0.026 U
0.046 U
0.024 U
0.035 U
0.036 U
0.243
0.019 U
0.030 U
1.263
0.036 U
0.037 U
0.016 U
0.039 U
0.039 U
0.057
0.024 U
5.141
0.035 U
0.037 U
0.038 U
0.037 U
35.8477
0.080 U
0.033 U
0.050 U
0.049 U
1.330 B
0.054 U
7.416
0.029 J
0.032 U
0.032 U
0.026 U
0.095 U
0.044 U
0.894 U
0.020 U
0.036 U
0.019 U
0.028 U
0.028 U
0.406
0.015 U
0.024 U
0.442
0.028 U
0.029 U
0.012 U
0.031 U
0.041
0.032
0.019 U
9.534
0.028 U
0.030 U
0.018 J
0.053
21.0375
0.091
0.052
0.076
0.077
1.046
0.170
4.384
0.054
0.049
0.049
0.040
0.057
0.960
0.469
0.039
0.054
0.030
0.042
0.043
0.853
0.024
0.037
0.425
0.042
0.043
0.020
0.048
0.053
0.030
0.030
2.460
0.043
0.050
0.072
0.055
12.067
                                                                  VOLATILE ORGANICS MONITORING REMEDIATION     287

-------
 SAMPLE NIMUII
                  Table 4
Volatile Organic Compound Concentrations for
        lenax Downwind Air Samples
      190JT-0)  1MST-11  190IT-19 190JT-1J 1M1T-10 1M1T-M tซ09T-4I AVTWCI
                                                OOMHUCI
      DOUMrUD  DGHWIND  DOMWIN) OOOMltD DOMMDB OOHMHIMD DGVWK)
                 11-19-BJ 11-31
                                          kl-04-Bt  ll-OS-M  1}<04-*I
  MIDfT MB COWT1TOWTS ppb
  Vinyl Chlorite
  CBlerathvM
  Mthylm* aittrl
  Anton*
  Carton Olmilfldi
  Chlorofon
  1,3-Olchl
  l.l,l-TrlcfeloroซllMป*
       1.1)4    1.M4    S.1I4   0.04J 0, O.ltl    ).?4? B 0.99) * 1.041
       1.U9   11.013    1-04) 0 0.014 J. 0.440 H,  T.DO   0.04* 0 1.4T1
       3.49)    0.144    O.UI J 0.141   O.)0?  13>.40B ป li,ซ?0 ป JO.tTt

       0.04) D   0.093    0.419 0 0.09* U 0.003 N,  0.1*0 0 O.OM 0 O.IM
       0.044 0   0.010 D  0.4)1 0 0.0)0 0 O.Oli 0  0.1*4 D 0.019 0 0.1)9

       0.04) 0   0.0)) D  O.tlB 0 0.019 0 0.01? 0  0.190 0 0.019 0 O.IM
       0.0)9 D   1.4?*    O.iil U 0.040 D O.OM D  0.1*1 D 0.0)9 V O.M*
       0.401    0.404    l.lli   0.0*4   O.M)    l.ป)   0.300   O.MO
       0.000    0.09T 0  O.IM U 0.019 B 0.01B D  0.1)1 0 0.01B P O.OM
       O.M* 0   0.014 0  O.?ll U 0.0)4 D 0.01) U  0.11* 0 0.0)) B 0.11*
       O.OM B   0.011 D  0.1?) V 0.01B B 0.01? B  O.IU 0 0.01? B O.OM

       O.OM B   0.01ป 0  O.JJ1 D O.OM 0 0.01) B  0.170 B 0.01* B 0.113
       0.01) V   0.449    0.4M 0 0.411   0.074    0.9)1   0.314   O.Mt
       O.OM 0   0.01B D  O.IM 0 0.014 B 0.01) U  O.O90 • 0.014 B O.OM
       0.0)3 B   0.0)0 U  0.4)9 D 0.011 B 0.030 B  0.141 0 0.011 B 0.1O4
       0.1*4    1.119    1.4*1   0.1)1   0.119    1.4 ti   O.MO   1.0B4
       O.OM ป   0.01B    O.ill U O.OM B 0.01) U  0.170 B 0.01) B 0.11)
  tUHtr  0.040 B   0.009 0  O.S?4 D 0.03? 0 0.0)1 B  0.1?? B O.OM B O.IM
       0.017 B   0.009 D  0,343 B 0.011 B 0.011 B  0.074 B 0.011 9 O.OM
       0.0)1
                         0.01) 0  O.tll 0  0.0)9 D  0.01? 0  0. 1M 0  O.OJf. B  0. Ill
                         0.01> D  0.411 J  0.0)ป   0.01T D  O.IM B  O.OM V  0.110


                         0.0)1 D  11.119   0.91?   O.JO   4*.0)4   I.1T9   11. 1U
                         0.03) 0  0.14) 0  O.OM D  0.014 0  0.141 0  0.011 B  0.111
                                                       O.OM B  0.09)
                                                       0.01? 0  0.179
                                                       O.OM V  O.IM
                  10.41M  J0.4MJT
                                    I.10H1  l.lt*01  It}. 4*9  11.7119   47.W
  The on-site samples collected SO ft downwind of the well point
installation are summarized in Ibble 5, along with average concen-
trations for all of the compounds. The major contaminants and
average concentrations are shown below:
          Compound
Toluene
Methylene chloride
Benzene
Trichloroethane
1,1,1  trichloroethane
                                    Average
                            Concentration  (ppb)

                                      39.5
                                       4.3
                                       3.6
                                       3.4
                                       2.3
   Lesser amounts of several other compounds also were detected,
   Tenax samples were collected at the on-site location during the
controlled release program. These samples are listed in Thble 6 along
with each sample number, type, date,  location and concentrations
for all HSL compounds analyzed for. The major compounds
288     VOLATILE ORGANICS MONITORING REMEDIATION
                                                                                                      Table 5
                                                                                    Volatile Organic Compound Concentrations for
                                                                                             Umax On-site Air Samples
                                                                                                  I9O9T-U I90ปf-U lMtT-31 190W-M 19OM-41 1MVT-4T
                                                                                            OttlTE   QHlft  OMIT!  OMITt  OMUTl  OMITf  OWITT
                                                                                                        I1-1H1 l)-01-ซ* l)-04-*l l)-W-f) !>•ซ?-•)
                                                                                MI coMrritmm
 4 -Nttky 1- 1-vmUMM
 T* t ?•<ป lATซซtปMM
 1,1.1. l-Tปt racftl*/ซซtM
0.4M


ป.4?4
0.4)1
I. Ml
0.31)

0.3S3
o.ra
0.14?

4. JIB
0.1*9
0,1*4
0.149
a. it?

ll.U?
0.11?
0.1U
ป.m
0.111
O.IM
0.0*1
O.IM
0.904
o.4n
0.14*
111. Ml
0.11?
O.U?
9.MU
101. 449



O.M)
B O.OM B
0.0?9


B 0.01) I



0.110



• O.OM B

1.0W
B O.OU B
• 0.094 f
l.M*

• O.OJO t
B 0.01J ป
B 0.011 B
B 0.01) B
V O.OM

B B.9M
• 0.034 t
* O.OM J,
B C.44I
1*.114ป 1



O.OM O.M4 9,
O.UO O.lOl 9
0.049 f 4.M4






O.OM B 0.ป>



O.OJ1 B O.OT1 V

O.OM B 0. MO

O.OM B 0.04) B
0,0*1 I.4C9

0.0U B O.M* B
O.OIt I O.OJI 9
OXM ซ :-.oปe c
0,OH • O.OM B
0.011 * O.OM

O.O41 15-491
0,081 V 0.041 t
O.U) 1 0,09? B
O.OM B ),•ป
I. Mart )?,9M) 1



9.444 10.0M
14.MI 1,OM
).)ป l.MT






*. 11) I. IB)



O.M* B O.M1

0.9)1 l.M)

O.M1 B 0.499
ป.?ซ 1.04V

0-)ป B 0,1't
0.1)8 B 0.14)
U. M* D 0.411
O.U) 0.411
0.400 0.1**

)!.)ซ M.441
0.111 B O.M)
0.1M B O.M1
1.1U O.BO?
n.mป 90.TU9

* 9.1*1 B
* o.in B
B 3.1*1 I
* O.MI f
J4.ll)
B O.IM *
B O.IM f
B 0.1MB



9. lit B



B 0.1U B
B 0.1MB
O.IU •

B O.IU B
0.19* •

B 0.144 C
B 0.0*1 *
• O.U1 B
B 0.153 •
J 0.09)9

1.440
B O.U* 9
B 0.147 B
0.144 9
M.*U*

..ป.
O.M)
ซ.HI
1.1*
i.m
ป.ป
*.ป
ซ.uป
ซ.!*
I.W

i.m
0.1U
O.M
C.1H
(.ซ
•.M
I.W
o.u>
A.W
1.H4
o.m
B.M1
>.m
o.m
ป.w
O.UI
O.UI
n.m
•.HI

i.m
W.1U
                                                             detected and average concentrations are shown below:
                                                                                   Compound
Toluene
Carbon  disulfide
Methylene chloride
1,1,1-trichloroethene
Acetone
Benzene
Trichloroethene
         Average
Concentration  (ppb)

          227
          191
            26
            12
             9
             5
             3
                                                                Several smaller quantities of numerous compounds also were
                                                              detected.
                                                                The average concentrations of upwind samples are compared to
                                                              average concentrations of samples taken downwind and samples
                                                              taken on-site both in the absence of and during controlled release

-------
              Table 6
Volatile Organic Compound Concentrations
For Tenax Controlled Release Air Samples
                  Table 7
Average Volatile Organic Compound Concentrations
    For Tenax Controlled Release Air Samples
SAMPLE KUKBES 1905T-28 1905T-34 1905T-38 1905T-40 AVERAGE
CONTROLLED
SAHPLE TYPE CS HP-4-C CR W-2-G CR HP-3-0 O> HP-4-0 RELEASE
(BACKUP-27) SAMPLE
SAMPLE DATE 12-O4-85 12-05-85 11-05-85 12-05-85 COHCEHTRATIOHS

AMBIENT AIR CONSTITUENTS ppb Dpb ppb ppb ppb

Chloraiethuie 0.600 D 0.685 D 0.685 0 0.800 0 0.693
BroraeUune 0.322 0 0.368 U 0.368 0 0.430 0 0.372
Vinyl Chloride 0.489 0 0.559 U 0.559 D 0.653 D 0.565
ChloroeUune 0.474 D 0.542 11 0.542 U 0.885 0.611
Methyleoe Chloride 10.443 J 39.509 B 21.812 B 30.729 B 25.623
Acetone 33.179 N,J 0.602 D 0.602 D 0.702 D 8.771
Cufcon Dlsulfloe 38.968 229.558 > 229.558 > 267.818 > 191.476
1,1-Dlchloroetbene 0.315 D 1.118 0.252 J 0.589 0.569
1,1-Dlchloroetbue 0.309 0 0.177 J 0.600 0.412 D 0.375
truu-l,2-Dlchloroetbene 0.315 D 0.361 U 0.361 0 0.421 0 0.364
Chloroform 0.256 B 0.293 0 0.293 D 0.342 D 0.296
1,2-Dlchloroethaoe 0.309 U 0.353 U 0.424 5.770 1.714
2-Butaione 1.421 D 0.485 U 0.485 U 0.566 D 0.739
1,1,1-Trlchloroethue 1.444 U 28.821 6.288 11.310 11.966
Cuton Tetrichlorlde 0.199 D 0.227 0 0.227 U 0.265 0 0.230
Vinyl AccUU 0.355 ป 0.406 0 0.406 U 0.474 0 0.410
Braodlchloroaetheiie 0.187 D 0.213 D 0.213 0 0.249 U 0.216
1,2-Dlchloropropone 0.271 D 0.309 0 0.309 0 0.361 0 0.313
truป-l,3-Dlcilloroi>ropซIK 0.276 0 0.315 D 0.315 D 0.368 U 0.318
Trlchloroetbene 1.024 5.055 1.330 2.793 2.551
aiorodlbranethue 0.147 D 0.168 0 0.168 D 0.196 U 0.170
1,1,2-Trlcbloroethuie 0.229 D 0.262 D 0.262 0 0.306 D 0.26S
Benzene 0.587 J 3.670 1.924 12.531 4.678
cls-l,3-01chloropropttDe 0.276 0 0.315 D 0.315 D 0.368 0 0.318
2-Chloroethyl Vinyl Ether 0.287 D 0.328 D 0.330 U 0.385 D 0.333
Broeofon 0.121 D 0.138 D 0.138 D 0.161 D 0.140
2-Hwunoae 0.305 El 0.349 U 0.349 D 0.407 0 0.353
4-Hetliyl-2-I>enUnoDe 0.305 D 0.593 0.768 0.773 0.610
TetrKhloroethene 0.184 D 0.422 0.548 0.172 J 0.332
1,1,2,2-Tetrachlorocthuie 0.182 V 0.208 U 0.208 U 0.243 U 0.210
Toluene 5.976 311.135 368.050 225.763 227.731
CUorobenune 0.272 D 0.311 D 0.311 0 0.362 D 0.314
Ethyl Beniene 0.288 D 0.362 0.527 0.384 D 0.390
Styrene 0.294 U 0.504 0.336 U 0.392 D 0.381
TOU1 Xylenel 0.288 D 2.799 1.548 0.384 D 1.255
TOTAL 100.899 631.519 641.412 568.760 485.648

in Ikbles 7, 8 and 9, respectively. Downwind concentrations of ace-
tone, carbon disulfide and toluene are generally 3 to 10 times higher
than upwind samples. Average concentrations of carbon disulfide,
acetone, toluene and methylene chloride in on-site samples taken
in the absence of controlled release are generally four to five times
PARAMETERS
Chloronettiane
Brooomethane
Vinyl Chloride
Chloroe thane
Methylene Chloride
Acetone
Carbon Disulfide
1 , 1-Dlchloroethene
1 , 1-Dlchloroethane

trans-l,2-Dichloroethene
Chloroform
1 , 2-Dichloroethane
2-Butanone
1,1, 1-Trichloroethane

Carbon Tetrachlorlde
Vinyl Acetate
Bromodlchlorone thane

1 , 2-Dich loropropane
trans-1 ,3-Dlchloropropene
Trlchloroetbene

ChlorodlbromoBethane
1,1, 2-Trichloroethane
Benzene

cls-1 ,3-Dlchloropropene
2-Chloroethyl Vinyl Ether
Broaoforo

2-Hexanone
4-Methyl-2-pentanone

Tetrachloroethene
1,1,2 , 2-Tetrachloroethane
Toluene
Chlorobenzene

Ethyl Benzene
Styrene
Total Xylenes
TOTAL
AVERAGE
UPWIND
(ppb)
0.091
0.052
0.076
0.077
1.046
0.170
4.384
0.054
0.049

0.049
0.040
0.057
0.960
0.469

0.039
0.054
0.030

0.042
0.043
0.853

0.024
0.037
0.425

0.042
0.043
0.020

0.048
0.053

0.030
0.030
2.460
0.043

0.050
0.072
0.055
12.024
AVERAGE AVERAGE
CONTROLLED CONTROLLED
RELEASE RELEASE
SAMPLE SAMPLES LESS
CONCENTRATIONS UPWIND
(ppb) (ppb)
0.693 0.602
0.372
0.565
0.611
25.623
8.771
191.476
0.569
0.375

0.364
0.296
1.714
0.739
11.966

0.230
0.410
0.216

0.313
0.318
2.551

0.170
0.265
4.678

0.318
0.330
0.140

0.353
0.610

0.332
0.210
227.731
0.314

0.390
0.381
1.255
485.315
0.320
0.489
0.533
24.577
8.601
187.091
0.515
0.325

0.316
0.255
1.657
-0.221
11.497

0.191
0.356
0.186

0.270
0.276
1.698

0.146
0.228
4.253

0.277
0.287
0.119

0.305
0.557

0.301
0.181
225.271
0.271

0.340
0.309
1.200
473.291
                                                 VOLATILE ORGANICS MONITORING REMEDIATION    289

-------
higher than upwind concentrations. In samples collected during the
controlled release program, concentrations are 8 to 100 times higher.
Other major compounds (1,1,1 trichloroethane, benzene and trich-
loroethene)  were detected at concentrations 4 to 20 times greater
than upwind sampling reported.

                               Table 8
          Average Volatile Organic  Compound Concentrations
                   For Tenax Downwind Air Samples

   PARAMETERS                  AVERAGE     AVERAGE     AVERAGE
                              UPHIND      DOWNWIND    DOWNWIND LESS
                              (ppb)       (ppb)       UPWIND
                                                      CONCENTRATIONS
                                                      (ppb)
                                0.091       0.205       0.114
                                                                           Table 9
                                                      Average Volatile Organic Compound Concentrations
                                                                For Tenax On-slte Air Samples
   Chloromethane

   Browne thane

   Vinyl  Chloride

   Chloroethane

   Methylene Chloride

   Acetone

   Carbon Dlsulflde

   1,1-Dlchloroethene

   1,1-Dichloroethane
 0.052       0.113       0.061

 0.076       0.169       0.092

 0.077       0.163       0.085

 1.046       1.603       0.557


 0.170       3.700       2.529

 4.384      16.238      11.854

 0.054       0.107       0.054

 0.049       0.112       0.063
   trans-l,2-Dlchloroetbene      0.049       0.108       0.060

   Chloroform                    0.040       0.069       0.048

   1,2-Dlchloroethane            0.057       0.108       0.051

   2-Butonone                    0.960       0.440      -0.520

   1,1,1-Trlchloroethane         0.469       0.669       0.200

   Carbon Tetrachlorlde          0.039       0.067       0.028

   Vinyl Acetate                 0.054       0.121       0.067

   BroBodlchloroKtbane          0.030       0.065       0.035

   1,2-Dlchloropropane           0.042       0.093       0.051

   trans-l,3-Dlchloropropene     0.043       0.095       0.052

   Trlchloroethene               0.853       0.287      -0.565

   ChlorodlbroaoBethane          0.024       0.051       0.028

   1,1/2-Trlchloroethane         0.037       0.081       0.043

   Benzene                       0.425       0.845       0.419

   cls-l,3-Dlchloropropene       0.042       0.093       0.051

   2-Chloroethyl Vinyl Ether     0.043       0.126       0.083

   Broaofon                    0.020       0.043       0.023

   2-Hexanone                    0.048       0.105       0.058

   4-Kethyl-2-pentanone          0.053       0.084       0.031

   Tetrachloroethene             0.030       0.071


   1,1,2,2-Tetrachloroethane     Q.030       0.0fi4

   Toluene
   Chlorobenzene

   Ethyl Benzene

   Styrene

   Total Xylenes


            TOTAL
                        0.041


                        0.034


 2.460      11.789       9.329


 0.043      0.094       0.051


 0.050      0.072       0.022


 0.072      0.100       0.028

 0.055      0.107       0.053


12.024    37.0815      25.058
                                                PARAMETERS
                           AVERAGE     AVERAGE     AVERAGE
                           UPWIND      ONSITE      OHSITE
                           (ppb)        (ppb)        LESS OPWIHD

                                                   COWCEHTRATIOHS
Chloroaethwe

BroaoM thane

Vinyl Chloride

Cbloroethane

Hethylene Chloride

Acetone

Carbon DlnilHde
Chlorobenzene

Ethyl Benzene

Styrene

Total Xylenes

          TOTAL
                             0.091       0.435       0.345

                             0.052       0.234       0.182

                             0.076       0.355       0.279

                             0.077       0.344       0.267

                             1.046       4.341       3.295

                             0.170       2.716       2.546


                             4.384        1.972      -2.413


l,l-Dlchloroethenซ           0.054       0.229       0.175

1,1-Olchloroethane           0.049       0.233       0.184

trani-l,2-Dlchloroetbeoซ      0.049       0.229       0.180

Chloroform                   0.040       0.186       0.145


1,2-Dlchloroethan*           0.057       0.224       0.167


2-Butanone                   0.960       0.777      -0.183


1,1,1-Trlchloroethane         0.469       2.344       1.875

Carbon Tetracnlorlde         0.039       0.144       0.106


Vinyl Acetate                O.OS4        0.258       0.204


BroBodlchloroaethane         0.030      0.136       0.106


1,2-Dlchloropropane          0.042       0.196       0.154


trans-l,3-Dlchloropropซm     0.043       0.200       0.157


Trlcbloroethene              0.853        3.444       2.591


ChlorodlbroBOBethane         0.024       0.107       0.083


1,1,2-Trlchloroethaoe         0.037       0.166       0.129


Benzene                      0.425        3.644       3.218


cls-l,3-Dlchloropropene       0.042       0.200       0.158


2-Chloroethyl Vinyl Ether     0.043       0.207       0.164


Broeufor*                    0.020       0.088       0.067


2-Hexanone                   0.048       0.222       0.174


4-Methyl-2-pentanone         0.053       0.344       0.291


Tetrachloroethene             0.030       0.222       0.192


1,1,2,2-Tetrachloroethane     0.030       0.132       0.103
Toluene                      2.460      39.499      37.039
                            0.043       0.197       0.155

                            0.050       0.506       0.456

                            0.072       0.258       0.186

                            0.055       1.973       1.919

                           12.067      66.556      54.489
 290     VOLATILE OROANICS MONITORING REMEDIATION

-------
 High Volume Filter Sampling

   Twenty high volume filters were collected and analyzed for HSL
 metals, total suspended particulates (TSP) and cyanide. The results
 listed in Table  10 show  that the field blanks had the following
 average concentrations of these major compounds:
                                                               Table 10
                                             High Volume Blank Filter Summary Lowry Landfill
                                                      Phase I Remedial Investigation
     Metal



     Aluminum

     Cadmium

     Magnesium

     Potassium

     Sodium

     Lead

     Zinc
Concentration  (yg/in2)



          29.3

          67.0

          28.7

          28.7

         345

           0.032

           0.4
   Table 11 lists the upwind samples and an average of these sam-
 ples for the HSL metals, cyanide and TSP concentrations.

   The Primary TSP National Ambient Air Quality  Standard
 (NAAQS) of 260 /tg/l/m3 (maximum 24-hr concentration not to
 be exceeded more than once per year) was established to protect
 the public health. The Secondary TSP NAAQS of 150 /tg/l/m3
 (maximum 24-hr concentration not be exceeded more than once
 per year) was established to protect the public welfare  from any
 known or anticipated adverse effects. These standards for particu-
 late matter are applicable only when 24-hr TSP concentrations have
 been measured by the U.S. EPA reference method or an equivalent
 method.
   The TSP method used during this investigation followed the refer-
 ence method guidelines with the exception that 8- to 12-hr sample
 periods were used instead  of  the nominal 24-hour period. The
 objective of the air quality TSP sampling was to measure particu-
 late, cyanide and metal concentrations during normal daily activi-
 ties at the landfill site. Sampling was performed during daylight
 hours when activities were highest. Consequently, the TSP upwind
 concentrations as shown in Table 11 range from 47 to 299 /ig/I/m3.
 The average TSP was 187 /tg/1/m3 for the seven upwind samples
 collected during this investigation. The Primary afNAAQS of 260
 /tg/l/m3 was exceeded twice, and the Secondary NAAQS of 150
 /tg/l/m3 was exceeded four times. 55ฎ  The downwind TSP samples
 shown in Table 12 averaged 325 /tg/l/m3 with a range of 112 to 643
 /tg/l/m3. Three of the samples were greater than the 260 /tg/l/m3
 guideline of the primary TSP NAAQS and four were greater than
 the Secondary  TSP  NAAQS of 150 /tg/l/m3.

  As shown in Table 13, the average downwind TSP concentration
of 326 /tg/l/m3 is considerably higher than the average upwind
concentration of  187 /tg/l/m3.

  The airborne HSL metals and cyanide concentrations as deter-
mined by CLP analysis  of the high-volume filters for the down-
SAMPLE HUHBER
DATE TAKEN
TYPE
VOLUME (H3)
FILTER HUMBER
TARE HEIGHT (CMS)
FINAL HEIGHTIGHSI
NET HEIGHT (CHS)
TSPIug/ซ3)

ALUMIHIUH
ANTIMONY
ARSENIC
BARIUM
BERYLLIUM
CALCIUM
CADMIUM
CHROMIUM
COBALT
COPPER
IRON
LEAD
HAGHESIUH
MANGANESE
MERCURY
NICKEL
POTASSIUM
SELENIUM
SILVER
SODIUM
THALLIUM
1905V-01
11-21-85
FIELD BLANK
HA
3920463
4.14170
4.14583
0.00413
NA
ug/lD2 UQ/B3
21.000 NA
0.800 U NA
0. 100 U HA
0.620 U NA
0.048 I NA
0.062 U HA
73.000 NA
0.100 U HA
0.330 U HA
0.250 U HA
0.600 U NA
0.023 0 HA
31.000 HA
0.200 U HA
0.003 U HA
0.370 U HA
29.200 U HA
0.053 U HA
0.160 D NA
378.000 NA
0.050 U NA
1905V-54
11-23-85
FIELD BLANK
HA
1090276
3.49208
3.51194
0.01986
HA
U,,ln2 u,/
21.000
0.800 U
0.100 U
0.850
0.830
0.062 U
38.000
0.270
0.330 U
0.250 U
2.200
0.023 0
15.000 U
0.200 U
0.003 U
0.370 U
29.200 D
0.053 U
0.160 U
255.000
O.OSO U
1905V-59
12-07-85
FIELD BLK
HA
3920468
4.11570
4.12069
0.00499
NA
'•3 ug/ln2 ug/ซ3
HA 46.000 HA
NA 0.800 U HA
HA 0.100 U HA
HA 0.620 U NA
HA 0.048 U NA
NA 0.062 U NA
NA 90. 000 HA
NA 0.100 U HA
NA 0.330 U HA
NA 0.250 U HA
NA 2.400 HA
HA 0.050 HA
HA 40.000 HA
HA 0.200 U HA
HA 0.003 U HA
HA 0.370 0 HA
HA 29.200 U HA
HA 0.053 D HA
HA 0.160 U NA
NA 403.000 NA
HA 0.050 U HA
AVERAGE
BLANK
CONCENTR





NA
ug/ln2
29.333
0.800
0.100
0.697
0.309
0.062
67.000
0.157
0.330
0.250
1.733
0.032
28.667
0.200
0.003
0.370
29.200
0.053
0.160
345.333
0.050


ATICt






ug/
NA
NA
HA
HA
HA
HA
HA
NA
HA
HA
NA
HA
NA
HA
NA
HA
NA
NA
NA
HA
HA
                                     TIN



                                     VANADIUM



                                     ZINC
0.450 0 HA



0.330 U HA



0.280  NA
0.450 D


0.330 D



0.680
HA    0.450 D KA



HA    0.330 U HA



KA    0.250  HA



NA    0.170 U HA
0.450  HA


0.330  HA



0.403  HA
                                                              VOLATILE ORGANICS MONITORING REMEDIATION    291

-------
                                                        Table 11
                                       High Volume Sample Filler Summary al Upwind Site
SAMPLE NUMBER 1905V-51
DATE TAKEN 11-19-85
TYPE UPWIND
VOLUME(M.l) 338.9
FILTER NUMBER 39204*7
TARE WEIGHT(GMS) 4.17282
FINAL WEICHT(CMS) 4.19120
NET WEICHT(GMS) 0.01838
TSP(ug/n3) 54.2342
ug/ln2 ug/>3
ALUMINIUM
ANTIMONY
ARSENIC
BARIUM
BERYLLIUM
CALCIUM
CADMIUM
CHROMIUM
COBALT
COPPER
IRON
LEAD
MAGNESIUM
MANGANESE
MERCURY
NICKEL
POTASSIUM
SELENIUM
SILVER
SODIUM
THALLIUM
TIN
VANADIUM
ZINC
CYANIDE
20.000
0.800 II
0.100 U
0.620 U
0.048 U
0.062 U
64.000
0.100 U
0.330 U
0.6(0 U
0.600 U
0.0(3
17.000
0.200 U
0.006
0.370 U
29.200 U
0.033 U
0.160 U
326.000
0.030 U
0.450 U
0.330 U
0.270
0.170 I'
3.7K
0.149 U
0.019 U
0.115 U
0.009 U
0.012 U
11.897
0.014 U
0.061 U
0.126 U
0.112 U
0.013
3.160
0.037 U
0.001
0.064 U
5.428 U
0.010 U
0.030 U
60.602
0.004 U
0.084 U
0.0*1 U
0.050
0.03? U
1905V-00
11-21-85
UPWIND
325.3
3920465
4.19853
4.21396
0.01543
47.4039

u|/ln2 ug/>3
23.000
0.800 U
0.100 U
0.620 U
0.048 U
0.06? II
71.000
0.100 U
0.330 U
1.900
0.600 U
0.0(3
27.000
0.200 U
0.003 U
0.370 U
29.200 U
0.053 U
0.160 U
344.000
0.050 U
0.043 U
0.330 U
0.270
0.170 U
4.457
0.155
0.014
0.120
0.009
0.012
13.742
0.019
0.064
0.368
0.116
0.016
5.226
0.039
0.001
0.072
5. (37
0.010
0.031
66.581
0.010
0.009
0.0*4
0.052
0.033
1405V-02
11-22-83
UPWIND
441.898
1090278
3.49599
3.61683
0.12084
245.6*0
U|/ln2
28.000
0.800 U
0.100 U
1.400
0.067
0.0*2 U
65.000
0.130
0.330 U
0.820
12.000
0.630 S
16.0OO
0.250
0.003 U
0.370 U
29.200 U
0.033 U
0.160 U
244.000
0.050 II
0.430 U
0.330 U
1.000
0.170 U

Ul/ปl
3.3(6
0.102 U
0.013 U
0.174
0.004
0.008 D
8.173
0.017
0.042 U
0.105
1.517
0.0(1 S
2.044
0.012
0.000 U
0.047 U
1.740 U
0.00? U
0.020 U
11. (41
0.006 U
0.058 U
O.O42 U
0.128
0.02? U
1405V-04
12-03-85
UPWIND
404.61
104027]
5.51288
3.60753
0.094*3
231.073

1403V-05
12-04-85
UPWIND
4(4.448
1040271
3.4(250
3.622(4
0.14014
248.446
Ug/ln2 ug/Ml U|/ln7
20.000
0.800 U
0.100 U
0.700
0.048 U
0.0*2 U
49.000
0.100 U
0.310 U
0.25O U
2.700
0.350 S
16.000
0.200 U
0.003 U
0.170 U
24.200 U
0.530 U
0.1*0 0
231.000
0.030 U
0.430 U
0.130 U
0.3(0
0.170 U
1.07*
0.121 II
0.013 U
0.10*
0.00? U
0.010 II
7.536
0,015 U
0.051 D
O.OM U
0.413
0.034 S
2.4*1
0.031 U
0.000 U
0.037 U
4.441 U
0.062 U
0.023 U
3(.603
0.008 U
0.0*4 U
0.031 U
0.058
0.026 1'
21.000 U
0.800 U
0.100
0.720 U
O.O48 U
0.0*7
44.OOO U
0.100 U
0.310 U
0.2(0
3.700
0.720
13.000 U
0.700 U
0.001 U
0.170 U
24.200 V
0.031 U
0.1*0 U
242.000
0.030 0
0.430 0
0.1(0
0.430
0.170 U

Uf/>l
2.818 U
0.107 U
0.011
0.047 U
0.00* U
0.008
6.576 U
0.013 0
0.044 0
0.038
0.497
O.OM
2.011 D
0.027 U
0.000 U
0.030 U
1.414 0
0.007 U
0.021 U
12.47*
0.007 V
O.OM I)
0.031
0.0*0
0.071 U
1405T-0*
12-04-85
UPWIND
447.513
1040270
1.444(1
1.63040
0.13107
263.444
Ug/tn2
14.000
0.800 U
0.100 U
1.500
0.048 U
0.0*2 U
36.000
0.100 U
0.310 U
0.570
3.800 V
0.130
15.000 U
0.250
0.001 U
0.170 U
34.300
0.530 U
0.1*0 U
242.000
0.030 U
0.430 U
0.470
0.610
0.170 U

U|/*3
7.40*
0.101 U
0.013 U
0.190
0.00* U
0.008 U
7.041
0.013 U
0.04? U
0.07?
0.4(1 U
0.01*
1.899 U
0.032
0.000 U
O.O47 U
4.977
0.0*7 U
0.020 U
3*. 97*
0.006 D
0.037 U
0.0*0
0.08?
0.072 U
190SV-08
12-07-83
UPWIND
348.992
10402**
3.48*42
3.34*34
0.05942
171.694
Uf/ln2
16.0OO
0.8OO U
0.100 r
0.700
0.048 U
0.0*2 D
1.2.000
0.100 U
0.310 0
0.750 U
0.6OO U
0.0(1
13.000 U
0.700 0
0.001 0
0.370 U
74.200 I)
0.031 U
0.1*0 U
271.000
0.050 D
0.430 D
0.310 I'
1.800
0.170 1'
AVERAGE
UPWIND
COHCEHTRATIOK
187.494
uป/.}
2.888
0.14* V
0.018 U
0.17*
0.004 U
0.011 0
7.5(2
O.OU U
0.0*0 U
0.045
0.108
0.015
2.708
0.016
0.001
0.0*7 0
3.271 U
0.010 D
0.024 V
40.73*
0.004 0
O.M1 U
0.0*0 ซ
0.123
0.011 U
U(/ln2
21.000
0.800
0.100
0.894
0.031
0.0*2
5*. 571
0.104
0.110
0.479
3.429
0.22*
17.2*4
0.214
0.003
0.170
30.641
0.1(4
0.1*0
273.2**
0.050
0.342
0.157
0.689
0.1 7O
<*>•}
1.277
0.126
0.01!
0.11]
0.007
O.OM
8.464
O.OU
0.031
0.111
O.tt*
0.01?
2.7M
0.011
0.000
0.058
4.7H2
0.027

41.91
0.007
0.059
0.05S
O.M
0.076
wind samples (Table 12) show the following major metals and
concentrations:
  Downwind

Concentration
                              All other HSL metals were below the detection limits. As shown
                            in Table 13, a comparison of the average metals and cyanide con-
                            centrations was made for  the  blanks,  upwind  and downwind
                            samples. A comparison of these for the metals detected is as follows:
     Metal



     Aluminum

     Barium

     Cadmium

     Lead

     Magnesium

     Potassium

     Sodium

     Zinc
  The upwind filter samples listed  in Table 11 show that the
following metals were detected:
23
1.2
51
0.31
17
29.2
274
0.96
3.8
0.21
7.9
0.04
2.6
4.7
44.6
0.15
                                                         tu^ 'ป * J      lug/in
29
0.7
67
0.03
29.2
143
0.401
21
0.9
36
0.23
30.*
27S
0.67
1.2
0.11
..9
0.01
..8
41.9
0.11
23
1.2
51
0.31
24.2
274
0.9*
1.8
0.21
?.ป
0.04
4.7
44.*
0.15
     Metal


     Aluminum

     Barium

     Cadmium

     Lead

     Magnesium

     Potassium

     Sodium

     Zinc
    Upwind

Concentration

(yg/inj)
   21

    0.9

   56

    0.23

   17

   30.6

 275

    0.67
 3.2

 0.13

 8.9

 0.03

 2.8

 4.8

43.9

 0.11
                           Sodltui

                           Zinc
                           All other  HSL  metals and cyanide were below the reported
                         detection limits.
                           The airborne heavy metal concentrations at the downwind site
                         were slightly higher for lead, aluminum, barium and zinc No U.S.
                         EPA or state standards currently exist for any airborne metals except
                         lead. The Primary and Secondary NAAQS for lead and its com-
                         pounds, measured as  elemental lead by the reference method or
                         an equivalent method, are each 1.5 pg/ms, maximum arithmetic
                         mean averaged over a calendar quarter. Although the filter samples
                         were collected and analyzed by the reference method, they were not
                         collected over a calendar quarter and thus are not directly corn-
                         parable  to the standards mentioned.
                           When indirectly compared, however, the downwind average con-
                         centration of 0.04 pg/I/m3, which is slightly  higher than the
                         upwind average concentration of 0.03 /ig/l/m3, is substantially less
                         than the Primary and Secondary NAAQS of 1.5 jtg/1/rn3. Com-
                         pared to the National Air Monitoring (NAM) Network average lead
                         values of 0.1 to 1.0 /tg/l/m3 for rural areas and 0.5 to 10.0 pg/l/m3
                         for urban areas, the upwind and downwind lead data show this area
                         to be in the low rural range.
292    VOLATILE ORGANICS MONITORING REMEDIATION

-------
                                                                              Table 12
                                                        High Volume Sample Summary at Downwind Site
                    SAMPLE NUMBER     1905V-52


                    DATE TAKEN       11-21-85


                    TYPE            DOHNHNIND


                    VOLUME(ป3>         582.9


                    FILTER NUMBER     3920464


                    TARE HEIGHTIGHS)  4.09579


                    FINAL HEIGHT IMS)  4.16140


                    NET HEIGHTIGHS)   0.06561
1905V-53
11-22-85
DOWNWIND
440.310
1090277
3.52626
3.59000
0.06374
1905V-55
12-03-85
DOHNHIND
542.923
1090274
3.50116
3.58440
0.08324
1905V-56
12-04-85
DOHNHIND
421.975
1090272
3.52338
3.67135
0.14797
1905T-57
12-05-85
DOHNHIND
298.509
1090269
3.50319
3.66712
0. 16393
1905V-60
12-07-85
DOHNHIND
260.317
1090265
3.50587
3.67328
0.16741
AVERAGE
DOHNHIND
CONCENTRATIONS





                    TSP(ug/>3)
                                   ug/ln2    ug/ป3    ug/ln2   ug/ป3    ug/ln2   ug/ซ3    ug/ln2   ug/ซ3    ug/ln2    ug/ซ3    ug/ln2    ug/>3    ug/ln2   ug/ซ3


                    ALUMINIUM         22.000    2.378    22.000     3.148   23.000    2.669   26.000    3.882   28.000    5.909    21.000    5.082    23.667    3.845


                    ANTIMONY           0.800 U  0.086 U  0.800 U   0.114 U   0.800 U   0.093 U   0.800 U  0.119 U  0.800 U  0.169 U  0.800 U  0.194 1)  0.800    0.129


                    ARSENIC           0.100 U  0.011 U  0.100 U   0.014 U   0.100 U   0.012 U   0.100 U  0.015 U  0.100 U  0.021 U  0.100 U  0.024 U  0.100    0.016


                    BARIUM            0.620 U  0.067 U  1.600     0.229    1.100    0.128    1.300    0.194    2.000    0.422    0.780    0.189    1.233    0.205


                    BERYLLIUM          0.048 U  0.005 U  0.048 U   0.007 U   0.048 U   0.006 U   0.067    0.010    0.050    0.011    0.048 U  0.012 U  0.052    0.008


                    CALCIUM           0.062 U  0.007 U  0.062 U   0.009 U   0.062 U   0.007 U   0.062 U  0.009 U  0.062 U  0.013 U  0.062 U  0.015 U  0.062    0.010


                    CADMIUM          63.000    6.809    56.000     8.013   44.000    5.106   54.000    8.062   53.000    11.186    36.000    8.712    51.000    7.981


                    CHROMIUM           0.100 U  0.011 U  0.100 II   0.014 U   0.100 U   0.012 U   0.100 U  0.015 U  0.170    0.036    0.100 U  0.024 U  0.112    0.019


                    COBALT            0.330 U  0.036 II  0.330 U   0.047 U   0.330 II   0.038 U   0.330 U  0.049 U  0.330 U  0.070 U  0.330 U  0.080 U  0.330    0.053


                    COPPER            6.100    0.659    0.980,    0.140    3.100    0.360    J.100    0.314    2.200    0.464    0.400    0.097    2.480    0.339


                    IRON              2.300    0.249    1.600     0.229    5.000    0.580    6.800    1.015    7.100    1.498    2.200    0.532    4.167    0.684


                    LEAD              0.700 S  0.076 S  0.520     0.074    0.270    0.031    0.220    0.033    0.083    0.018    0.067    0.016    0.310    0.041


                    MAGNESIUM         24.000    2.594    15.000 II   2.146 U  15.000 U   1.741 u  15.000 U  2.239 U 15.000 U  3.166 U  15.000 U  3.630 U  16.500    2.586


                    MANGANESE          0.200    0.022    0.200 U   0.029 U   0.200 U   0.023 U   0.200 U  0.030 U  0.200 U  0.042 U  0.200 U  0.048 U  0.200    0.032


                    MERCURY           0.005    0.001    0.003 U   0.000 U   0.003 U   0.000 U   0.003 U  0.000 U  0.003 U  0.001 U  0.003 U  0.001 U  0.003    0.000


                    NICKEL            0.370 U  0.040 U  0.370 U   0.053 U   0.370 U   0.043 U   0.370 U  0.055 U  0.370 U  0.078 U  0.370 U  0.090 U  0.370    0.060


                    POTASSIUM         29.200 U  3.156 U  29.200 U   4.178 U  29.200 U   3.388 U  29.200 U  4.359 U 29.200 U  6.163 U  29.200 U  7.067 U  29.200    4.719


                    SELENIUM           0.053 U  0.006 U  0.053 U   0.008 U   0.053 U   0.006 U   0.053 U  0.008 U  0.053 U  0.011 U  0.053 U  0.013 U  0.053    0.009


                    SILVER            0.170    0.018    0.160 U   0.023 U   0.160 U   0.019 U   0.160 U  0.024 U  0.160 U  0.034 U  0.160 U  0.039 U  0.162    0.026


                    SODIUM           336.000    36.315   248.000    35.484   232.000    26.921   210.000    31.353   351.000    74.078   264.000    63.891   273.500    44.674


                    THALLIUM           0.050 U  0.005 U   0.050 U   0.007 U   0.050 U   0.006 U   0.050 U  0.007 U  0.050 U  0.011 U  0.050 U  0.012 U  0.050    0.008


                    TIN              0.450 U  0.049 U  0.450 I)   0.064 U   0.450 U   0.052 U   0.450 U  0.067 U  0.450 U  0.095 U  0.450 U  0.109 U  0.450     0.073


                    VANADIUM           0.330 U  0.036 U  0.330 U   0.047 U   0.330 U   0.038 U   0.330 U  0.049 U  0.330 U  0.070 U  0.330 U  0.080 U  0.330    0.053


                    ZIKC              0.680    0.073     1.400     0.200    0.950    0.110    1.000    0.149    1.300    0.274    0.430    0.104    0.960    0.152


                    CYANIDE           0.170 U  0.018 U  0.170 U   0.024 U   0.170 U   0.020 U   0.170 0  0.025 0  0.170 D  0.036 U  0.170 U  0.041 U   0.170    0.037
Polyurethane Foam (PUF)  Sampling
   Polyurethane foam (PUF) sampling media were used to  collect
upwind and downwind air quality samples. The PUF samples were
analyzed for  semivolatiles and pesticides/PCBs under the CLP.
   The batch blank (1905P-02) and the field blank (1905P-03) had
29  and   16  /tg/1  per   polyurethane   foam  (^g/l/PUF)  of
     BIS(2-ethylhexyl)phthalate, respectively. There were no other con-
     taminants reported. All of the samples were contaminated with trace
     amounts of this semivolatile compound.
        The  following  comparison  of  downwind  and upwind  PUF
     samples indicates that the majority of the compounds are slightly
     higher  at the  downwind location:

VOLATILE ORGANICS MONITORING REMEDIATION      293

-------


Phenol
2-methylphenol
4-methylphenol
Naphthalene
2-methylnซpthซlene
Acenapthalene
Diet hylpht ha late
Fluorene
Phenanthrene
Anthracene
Di-N-butylph thai ate
Bis (2-ethylhexyl)phtha late

Upwind (ppb| Downwind Ippb)

0.01 0.06
0 .005 0.012
0.003 0.012
0.02 0.011
0.18 0.49
not detected 0.002
0.002 0.001
not detected 0.002
not detected 0.00}
0.001 not detected
not detected 0.019
0.02 0.04

Table 13
Comparison of Average
High Volume Blank, Upwind
And Downwind Samples
AVERAGE

BLANK

CONCENTRATIONS

TSP(ug/ซ3) NA

ug/ln? ug/ซ3

ALUMINIUM 29.333 NA

umxufi o.soo NA
ARSENIC 0.100 NA
BARIW 0.69? NA
BERYLLIUM 0.309 NA
CALCIOM 0.062 NA

CADMIUM 67.000 NA

CXROKIDH 0.157 NA
COBALT 0.330 NA

COPPER O.J 50 NA
IRON 1.7)1 NA

LEAD 0.032 KA
MAGNESIUM 26.667 NA
MANGANESE 0.200 NA
MERCURY 0.003 NA

NICKEL 0.370 NA

POTASSIUM 29.200 NA
SELENIUM 0.05} NA
SILVER 0.160 NA
SOOIBM 345.333 NA
THALLIUM 0.050 NA
TIN 0.450 NA

VANADIUM 0.330 NA

ZINC 0.403 NA
CYANIDE 0.170 NA
AVERAGE AVERAGE

WIND DOUHHIND

CONCENTRATIONS CONCENTRATIONS

187.494 325.59)

ug/ln2 UQ/B) UQ/1D2 U9/B3

21.000 3.271 2). 667 3.845

0.100 0.126 0.600 0.129
0.100 0.016 0.100 0.016
0.194 0.134 1.23) 0.205
0.051 0.006 0.052 0.00*
0.062 0.010 0.062 0.010

56.571 8.964 91.000 7. Ml

0.104 0.016 0.112 0.019
0.1)0 0.052 0.330 0.053

0.679 0.113 2.480 0.139
3.429 0.467 4.167 0.664

0.226 0.032 0.310 0.041
17.286 J.788 16.XX> 2.586
0.214 0.03) 0.200 0.032
0.003 0.001 0.00) 0.001

0.370 0.016 0.370 0.060

30.613 ..761 29.200 4.719
0.189 0.027 0.053 0.009
0.160 0.015 0.162 0.026
275,286 43.912 273.500 44.674
0.050 0.006 0.050 0.008
0.392 0.060 0.450 0.07)

0.357 0.055 0.330 0.053

0.689 0.108 0.960 0.15}
0.170 0.027 0.170 0.027
In addition, all PUF samples were analyzed for HSL pesticides
and PCBs. None of the samples (including the blanks) had con-
centrations over the detection limits of the methodology as sped-
Tied in the U.S. EPA guidelines.
Samples 1905P-54, 1905P-55 and 1905P-56 were spiked in the field
by U.S. EPA contract laboratory personnel with 1 ml of 1.07 mg//d
Aldrin and 1.0 mg/^l PP Methoxychlor. Other than trace amounts
of Bis(2-ethylhexyl)phthalate, nothing was detected in these spiked
samples. It is not known at this time why these samples read clean.
Carbon Molecular Sieve (CMS) Sampling
Carbon molecular sieve (CMS) media were used to collect air
quality samples at the upwind, downwind and on-site locations for
analysis of VOCs. The field blanks reported clean except for trace
amounts of toluene. Average concentrations for major compounds
detected in upwind samples are shown at 1.46 ppb for methylene
chloride, 0.42 ppb for acetone and 0.83 ppb for carbon disulfide.
Trace amounts of trichloroethene and toluene also were present in
the upwind samples.
The following list shows the average concentrations of major
VOCs found in downwind samples:
Downwind
Compound Concentration (ppb)

Carbon disulfide 1.1

Methylene chloride 0.21

Acetone 0.11

1,1 dichloroethene 0.85

1,1,1 trichloroethene 0.06

Benzene 0.09

4-methyl-2-2pentanone 0.07
Toluene 15.5
Ethyl benzene 0.07

Styrene 0.06

The high toluene average concentration was influenced by a single
sample (1905C-I4) that showed concentrations greater than 50.4 ppb
(7,400 ng/tube). The downwind toluene average without this sample
is 9.2 ppb. The remaining downwind samples had relatively low
toluene concentrations.
The CMS samples collected on-site in the absence of controlled
release showed the following average concentrations for listed com-
pounds:
Onsite
Compound Concentration (ppb)

Carbon disulfide 1.95

Methylene chloride 0.46

Chloroethane 0.44
Acetone 0.61
1,1 dichloroethene 0.3
1,1,1 trichloroethene 0.21

Benzene 0.36

Toluene 1 . 35

Styrene 0.27
294    VOLATILE OROANICS MONITORING REMEDIATION

-------
  Of the samples collected during the controlled release program,
only carbon disulfide samples showed a substantial average con-
centration (5.6 ppb). A comparison of major compounds in CMS
samples from all sampling events is given in Table  14.

                            Table 14
                    CMS Sample Comparison
       Compound




   Orbon dlwlflde

   Hethylene chloride

   Chloroethซnซ

   Acetone

   1,1 dlchloroethene

   Ethyl benzene

   Styrene

   Toluene

   Benzene

   1,1,1 trlchloroethene

   4-ซethyl-2-pentซnone
Upulnd




 0.8}


 1.46
                                            Reguler

                                            Onilte
                               Control led

                                teleiie
 1.1


0.21




0.11


0.8S


0.07


0.06


15.5'


0.09


0.06


0.07
                      0.46


                      0.61


                       0.3
                      0.27


                      0.16


                      0.21
   'Reflect! one tuple (1905C-14) reporting ซreซter tlun 7400 ni/tube (9.2 ppb without thlซ

   suiple).

   The comparison shows that downwind samples have an average
 higher concentration than upwind samples for all compounds listed
 except methylene chloride and acetone. On-site sample concentra-
 tions in the absence of controlled release were higher than upwind
 samples for all compounds except methylene chloride and carbon
 disulfide.
  The presence of only one measurable compound, carbon disul-
fide, in the sample taken on-site under controlled release condi-
tions may be a result of insufficient sampling time, i.e., there may
not  have been enough sample volume  collected for adequate
amounts of other constituents  to collect in the tube.

CONCLUSIONS
  The short-term ambient air investigation, as previously discussed,
was undertaken during the winter months and produced a limited
ambient data base. Because of the insufficient number of sampling
locations and sampling duration, it cannot be certain the maximum
impacts were measured by the short-term  program. A more exten-
sive sampling network operated over a long period of time would
be required to generate a valid data base.
  Based on the short-term ambient data, however, we conclude that
there is some indication that downwind concentrations  of acetone,
carbon disulfide  and toluene are three to  10 times higher than
upwind concentrations. The samples collected on Tenax at the con-
trolled release site were generally 8  to 100 times higher than the
upwind air quality for these compounds. Other compounds, such
as 1,1,1-trichloroethane, benzene and TCE,  were also detected at
concentrations four to 20 times greater  than upwind sampling
reported.
  TSP levels appear to be much higher downwind them upwind.
The  daily activities on-site are probably the primary cause. It
appears that the volatile organic compound levels (ranging from
a few ppb to several hundred ppb downwind of the controlled release
sites) are relatively low, as compared to worker exposure threshold
level values (TLVs), however a risk analysis should be  completed
before any decisions are made.

REFERENCES
1. US. EPA Compendium of Methods for the Determination of Toxic
  Organic Compounds in Ambient Air, EPA-600/ 4-84-041,  April 1984.
2. U.S. EPA Quality Assurance Handbook for Air Pollution Measurement
  Systems, Volume II Ambient Specific Methods,  Section 2.0, EPA
  600/4-77-027a, April 2983.
3. US. EPA 40 CFR, Appendix B, as amended Dec. 6, 1982 (47 FR 54912).
                                                                 VOLATILE ORGANICS MONITORING REMEDIATION     295

-------
               Lessons  Learned  in  Remedial  Design  and  Cleanup
                                   Of  a Federal  Superfund  Site

                                              Thomas F. Maher, P.E.
                                      William F.  Cosulich  Associates,  P.C.
                                               Woodbury,  New York
                                          Michael W.  Mel aughlin,  J.D.
                                                   SCS Engineers
                                                  Reston, Virginia
                                                  Christine Beling
                                     U.S. Environmental Protection  Agency
                                               New York,  New York
ABSTRACT
  The  Krysowaty Superfund site was  one of the  first  NPL
hazardous waste excavation projects to be successfully completed.
The remedial action selected for the site was excavation and off-
site disposal of waste  and contaminated  soil. Because of the
commitment made by the Federal Government to the local com-
munity, this project was "fast-tracked," and design (including a
pre-design investigation) was completed within 4 months, and con-
struction was completed 6 months after design. However, a number
of special requirements and problems were encountered in under-
taking this project implementation approach, and in design and
construction itself.

The following will be discussed:

• The kinds of information needed to prepare design plans and
  specifications, as compared with the types of information typi-
  cally generated during the RI/FS. Information needed includes
  soil mechanics (for slope stability), limits of contamination (for
  plans and specifications and cost estimate) and shallow ground-
  water data (for dewatering purposes).
• Establishment of a framework  to make decisions early in the
  design process through the use  of explicit design criteria. The
  specific design criteria developed are presented.
• The use of performance versus design specifications. The project
  was designed differently from some recent Superfund projects,
  in that  a specific design was prepared for bid. This approach
  offered advantages over a more general performance specifi-
  cation approach,  and the advantages are discussed.
• The successful approach taken to estimating construction cost,
  including special steps which take into account the incremental
  costs of worker health  and safety.
• Special specifications developed to comply with the new RCRA
  off-site disposal regulations.
• Bid schedule, including strategy used to assure that contractor
  bids  would  enable the Government  to add  additional  exca-
  vation, transport, disposal and backfill  in accordance with
  equitable unit prices.

INTRODUCTION
  Krysowaty Farm, which is ranked 104th on the NPL, is located
in central New Jersey. The site is situated  on a 42-acre tract of
land in Hillsborough Township, Somerset County, near the Village
of Three Bridges. Disposal of chemical wastes at the site was
reported to have occurred between 1965 and 1970. An estimated
500 drums of paint and dye wastes and  unknown materials were
allegedly  dumped, crushed and buried at the site. In addition to
drums, other wastes including demolition debris, tires, automo-
biles, bulk waste, solvents, waste sludge and other material were
disposed at the site.
  Several studies were undertaken regarding the  nature of the
wastes buried at the site, environmental and public health impacts,
and surrounding geology and hydrology. This included a RI/FS
conducted under the direction of the U.S. EPA and completed in
March  1984.
  A wide range of contaminants was identified in soil and sedi-
ment on-site. There was not a consistent pattern of distribution
of the contaminants and each compound appeared to have a unique
distribution in the soil and sediment. Approximately 40 different
compounds were identified, including base neutral compounds,
volatile organic chemicals,  pesticides and PCBs.
  In addition to  soil and  sediment analysis, composite waste
samples were obtained from 20  drums uncovered during test pit
excavation. Many of the same compounds detected in the drums
and surrounding soils were also found  in seeps originating from
the  site and in the underlying groundwater.
  Based on topographic analysis, power auger borings and mag-
netometer survey performed during the RI/FS, the aerial extent
of buried waste material  was estimated to be approximately 0.5
acre and the average thickness of  the material was estimated to
be 5 ft. The recommended remedial action based on this informa-
tion and contained in the ROD was to excavate this area comprising
an estimated 4,000 yd3 of waste and contaminated soil including
the  first 6 in. of bedrock for off-site disposal.
  Field investigation  conducted as part of Preliminary Design
indicated that, based  on detailed  topographic and seismic data
obtained in the immediate area of the disposal site,  the waste area
depth was more extensive than originally estimated. In addition,
a review of earlier magnetometer information and soil boring data
indicated that the waste area may be more extensive aerially than
contained in the ROD, and  the total volume of waste specified to
be removed in the design was increased to  8,500 yd3.
  Because of the commitment made by the Federal government
to Hillsborough Township, this project was "fast-tracked" and
design was completed within 4 months. These steps included a Pre-
Design Investigation, preparation of contract plans and specifica-
tions and development of a detailed construction  cost estimate,
comprising  Preliminary,  Advance Final  and  Final Design
documents. As a result of recent amendments to the RCRA, special
specifications were developed for waste characterization and off-
site disposal to ensure compliance with the new regulations.
  The project was successfully bid and the contractor was selected
within 3 months and construction was completed within 3 months.
 296    SITE REMEDIATION

-------
In addition to excavation and off-site  disposal, construction
included in-place waste characterization to (1) provide necessary
information for acceptance of the waste material at a permitted
RCRA disposal facility and determine the limits of contamination
prior to full-scale construction.
  Although cleanup of the site was  successfully completed, a
number of special requirements and problems were encountered.
A discussion  of  these  requirements  and  problems  and their
resolution follows.

LIMITS OF CONTAMINATION AND
CONSTRUCTION  BID SCHEDULE.
  One primary objective of a RI/FS  is to define the extent of
contamination. In many investigations, sampling points are located
at the nodes of a grid network based on 50-ft or greater spacing.
Often  the limit of  contamination is  estimated  to be mid-way
between the last contaminated node and the first "clean" location.
Based on this method of contamination delineation, an estimate
of volume (in this case waste and soil) is calculated and a  cost per
cubic yard for excavation, removal and/or treatment is  used to
determine the cost  of remedial action, exclusive of  design and
contingencies.
  In  performing a relative  cost  comparison  analysis  among
remedial alternatives, which typically is performed in Feasibility
Studies, the actual amount of contamination often is not critical
in selecting the  most cost-effective option. However,  for  the
purposes of design and accurate cost construction estimation and
preparation of a Bid Schedule, it is important to know as precisely
as possible, the exact  limits and  quantity of  material to be
remediated.
  As a general example, Figure 1 illustrates a typical grid mid-point
delineation of contamination and a possible actual limit. The
volume of actual contamination in this situation is 70% greater
than that estimated  by assuming that the limit of contamination
is at the grid mid-point. The difference in volume between  the mid-
point and actual is about 4,700 yd3, which, at a cost of $300 yd3
for excavation, removal  and disposal, results in a  cost differential
of $1.4 million. Failure to determine the correct amount of  material
to be remediated generally results in an insufficient obligation of
funds for cleanup,  delays in the project because of the  need to
appropriate additions  funds  and  major construction  Change
Orders.
  Recognizing this possibility in the design of remedial action for
Krysowaty Farm and having no time, because of the "fast-track"
approach  taken with regard to this project to conduct sufficient
pre-design investigation (soil borings/test pits) to better define the
actual limits of contamination, an estimate was made of  the need
for possible additional excavation. Also provision was made for
test pit excavation by the contractor as part of construction prior
to commencing full-scale excavation. The possible need for addi-
tional excavation and removal, together with a more detailed cost
estimate for hazardous  construction  resulted in an  increase in
estimated  construction,  costs from $1.5 million as defined in the
RI/FS to nearly $3.8 million. This new  estimate was used to secure
an additional obligation of funds prior to construction and thereby
avoided project delays.
  In addition to pre-determining the need for additional construc-
tion and funds, the Bid Schedule was designed to provide maximum
flexibility for the U.S. Army Corps of Engineers (which was respon-
sible for design  and construction management) and U.S. EPA
(which was responsible for funding and oversight) to modify the
plan for excavation during construction as a result of field changes
without major Change Orders.
  This Bid Schedule, which is illustrated in Table 2, was designed
to provide a baseline bid based on the estimated quantities with
unit prices and an optional bid, also with unit prices and estimated
quantities, to handle additional construction as necessary. This sub-
divided item approach with options, as opposed to a lump sum
type of bid, proved successful in the project and allowed the Corps
                                ACTUAL LIMITS Of CONTAMINATION
                             LIMIT* OF CONTAMINATION DEFINED
                    I          IT GRID MID-POINT

                           Figtwe 1
         Comparison of Possible Actual Versus Typical Grid
              Mid-Point Delineation of Contamination
of Engineers and U.S. EPA to undertake additional required work
to remove contaminated soil and waste (including contaminated
marsh sediment downgradient of the site which was not in the
original design, without Change Orders).

DETAILED DESIGN SPECIFICATIONS
  In the initial design stages for the Krysowaty Farm site, there
was interest expressed by the COE that performance specifications
be used as the basis for design. It was thought that this approach
would result in less chance for error in design and therefore less
need for potential Change Orders. This would also allow the con-
tractor more freedom for innovative design and thereby possibly
reduce construction costs. However, the use of performance specifi-
cations can also result in unsatisfactory bids, project delays due
to unapprovable Shop Drawings and the need to possibly rebid
the project.
  As an example, the performance specification for a decontami-
nation pad may simply require that the pad not release contami-
nation; if there is  a release, cleanup would be the responsibility
of the contractor. This approach could result in irresponsible bids;
a contractor could, in order to keep construction costs low and
thereby be selected as the  lower bidder, after selection, submit a
Shop Drawing which shows only a simple plastic liner with runoff
to a buried drum. Such a system likely would be subject to failure
and would not be  approved. If the contractor chose not to offer
a suitable alternative at the same bid price, the project would need
to be rebid and the project would be delayed.
  To avoid this pitfall, it was decided to design the remedial plan
for Krysowaty Farm using detailed construction specifications
(dimensions, materials, specific facility  locations, etc.) so that all
construction would meet the intent of the project and all contrac-
tors would be bidding on  the same project.  In  order to provide
the opportunity to reduce costs, the contractor was offered as part
of a Value Engineering Specification in  the bid document, the
opportunity to propose alternatives to the design together with the
                                                                                                  SITE REMEDIATION     297

-------
                            Table 2
                       Krysowaty Farm Site
                     Hillsborough, New Jersey
                          Bid Schedule
                         EatUated
                         Quantity
                                         Unit
                                         trill
          All work eoaplot*
          except for the
          •undivided Itcu
          lletod belov.
          Include! Hobll-
          zatlon end Sit*
          Preparation; end
          De*obllUetlon
          Test Pit Excevetlon
                          Subdivided Itei
                                   pit
          Meat
          uid
          for
         Execution
         reparation
     Addl Ion... Wnt*
     be* at Ion antl
     Prcparacton for
     Diซpoซซl  (Optional)
     OruB •••oval end
     Preparation for
     Truuport

     Addltieul On.*
     I.WOT.II txad prซ-
     opซrซtloe for
     Truuport (Option*!)

     Off Sltt Truu-
     port uwl Dlspotul
                           2600 ton*  too
                           10 druu   orm

                           1100 conn  con
5ซ.   .Utlttioiu.1 Off Sltt    2600 toni  ton
     Truuport utml
         Dlapoeal (Optional)
         BleekMll
         Cloonre
                            10 tons  drun
                          3500 CT
     ti.   Additional Isckflll
         uul Cloeurv (Optional)  2000 cr
         TOTAL U10OTT
                                             l/plt-
                                             I	.
l/ton-
I	
                                             l/ten-
                                             I	
                                             I/or
                                             I	
                                            l/dr
                                            I	
                                            I/to
                                            I	
                                            l/dr
                                            I	
                                           t/ten-
                                           I	
                                       t/cj-
                                       I	
                                       t/cy-
                                       I	
attendant reduction in cost. As an incentive, the contractor would
receive 50% of the cost savings. The use of both detailed specifi-
cations and value engineering proved successful.

SPECIAL  SPECIFICATIONS FOR
OFF-SITE DISPOSAL
  Since this was one of the first, Superfund construction projects
conducted  under the then new RCRA hazardous waste off-site
disposal policy, special specifications were developed to ensure that
hazardous waste excavated as part of the project was transported
and disposed in conforrnance with the new RCRA policy. Detailed
provisions were included in the bid document which required the
contractor to ensure and provide evidence that the facility, at which
the hazardous waste was to be disposed, was in conforrnance with
all RCRA requirements. These requirements included the following:

• The facility must have been inspected by the appropriate state
  or Federal officials responsible for the RCRA or TSCA program
  within 6 months prior to receipt of wastes under this contract.
• The facility must not have any significant Class 1 RCRA vio-
  lations or other environmental  conditions that affect the satis-
  factory operation of the facility.
• Under limited circumstances, the U.S. EPA Regional Adminis-
  trator  may allow disposal of hazardous substances at a RCRA
  facility having significant RCRA violations or other environmen-
  tal conditions affecting satisfactory  operation,  providing that
  the facility owner or operator has entered into a consent order
  or decree to correct the  problems  and disposal  only occurs
  within the facility at a new or existing unit that is in compliance
  with RCRA requirements.
• If the facility is a land disposal facility, disposal must be in a
  unit meeting applicable RCRA minimum technical requirements
  (if RCRA waste is disposed) and include the use of a double liner
  system.
• Under limited circumstances (low waste toxicity, mobility and
  persistence), the U.S. EPA may approve the use of a single-lined
  land disposal unit for RCRA wastes where use  of such a unit
  adequately protects public health and the environment. If the
  Contractor proposes to use such a single-lined land disposal unit
  for RCRA wastes removed  from the site, the Contractor shall
  include evidence of U.S. EPA approval with its bid.
• The Contractor should contact the Regional Off-site Coordina-
  tor to determine the status of the disposal  facility and assure
  that the selected site meets  approval just prior  to waste exca-
  vation and disposal.

GEOTECHNICAL AND HYDROLOGIC INFORMATION
  Most RIs collect information of a scientific rather than an
engineering nature. RIs collect data through chemical sampling and
field studies to identify contaminants  of concern, exposure path-
ways and otherwise characterize the site to support or refute a scien-
tific hypothesis of the problem(s) presented by the site.
  One common criticism of the remedial  investigation  process is
that not enough "hard" engineering data are collected during the
process, thus requiring the design engineer to conduct additional
investigation or  make conservative design assumptions before
beginning to prepare plans and specifications for the remedial
action.
  It was evident  early in the Krysowaty Farm  Remedial Investi-
gation that excavation in  some  form  might emerge as a part of
the eventual remedial action.  To design a waste site excavation,
it is important to know:

• The mechanical properties (e.g., triaxial test, Atterberg limits,
  grain size) of the materials to be excavated to determine whether
  shoring is needed at the face of the excavation
• The water table elevation in the excavation area to determine
  whether dewatering is needed

  Numerous soil  borings were installed during the RI of the site,
but no soil samples were collected for geotechnical evaluation. The
cost to  obtain the needed geotechnical information during the
investigation would have been  negligible; however,  to mobilize for
a separate subsurface investigation during the remedial  action
design would have been significant.
  Because of time constraints and cost conservative assumptions
were  used in the design and  the bid  specifications, rather than
mobilize drill rigs and collect geotechnical data. These assumptions
included restrictions on the height of exposed excavation faces
(5 ft) and the steepness of grades (1:1 slopes) over that height.
  The excavation area at  Krysowaty  Farm was just uphill from
a spring which appeared to emerge from the contaminated fill area.
The proximity of a spring suggested that  portions  of the excava-
tion area  might have to be dewatered prior to or  in conjunction
with excavation. However, the  RI report did not mention the source
and path  of groundwater.
  Again, no time was available to perform  subsurface investigation
during design. The design assumed that  a subdrain installed in
bedrock  at  the top of  the fill area would intercept and divert
groundwater  from the excavation area. During construction,
however, it was found that groundwater did not flow near or above
the bedrock surface, and  other methods  were needed  to control
the spring during excavation.

EXPLICIT DESIGN CONTROL
  One of the most important early design steps was to develop
criteria for review and comment by the COE and other interested
government agencies.
  Although submission of preliminary  design criteria was not
required by the design contract, these criteria were prepared and
 298    SITE REMEDIATION

-------
                                 Table 3                                         submitted anyway. The short period of performance contemplated

                      Preliminary Design Criteria                              for design made it necessary to take every precaution to assure that

itifdin >ซi.p. cwiponnvi	fr.p.ied en ten™ ,M tmmt* •ป mim	        au involved parties were in agreement on the critical components


—— 'rt'" """"                   Xii'n.  -l"Kซ1. Tn"".,."1.;.1 "Iat.ri:;.       of the remedial action.

                                       Int^iJ.ted~..i.'t.n"*"n I^*lp.ri'>l'or".ปdfin>          Table 3 shows the preliminary design criteria initially proposed

                                       .u,.rfป,\i"t"ฐ.n?ฐth'.*"n?rฐ.™trudct'u'ri.' ri^Tr.!       for the project. As expected, these criteria generated considerable

                                       ..bruVno^proti.0,"""","  ฐlw™Z',       discussion among the involved parties.  Several adjustments were

                                       deb'"'."!"?, inciudedln'weeifie"!™"'"10"'       made in the criteria in response to comments. The final basis for

Quentltr of be.nted H.t.tl.1.                 lecord of D.cl.lon ep.clfled 4,000 cubic  y.rd.       dCSlgH  IS SUmmaHZed Ul thlS Table.
                                       for uc.r.tlon uid off.lt. dl.poe.l.   K.ther


                                       Srซ ruriEol"VH^.?r*^i       CONSTRUCTION COST ESTIMATE
                                       tecord of Decl.lon eppe.r.  to  undere.tl.ete            .      .       , „.,. -    ....           -            .    .   rr
                                       roi_. of ut.n.i .hicb .in  be eie.r.ted by.          The estimated $3.8 million cost of constructing the Krysowaty
                                        f.etor of  Jiut orer two, uu  t,SOO cubic y.rd.                                 .                             " „  .    ..      ,
                                       or 11,400 ton... b..i. for n.iu.tim bid..          Farm remedial action included:  excavation and off-site disposal

n.it. .f iป.r.tion                         i.cord of  D.ci.i.11 .pecified  ซc.Y.ti.n .ithin       of 6,800 yd3 of waste and contaminated soils (base bid); construc-
                                        "v..t. dl.po.Bl  .re."  down  to  bedrock, wd        .-            .     ป  ,  •    i          •    •       j      •
                                        rซ.eซi of  fir.t  .u  inch.,  of  bedrock.       lion of access roads, subdrain, decontamination and staging areas;
                                       Borltont.l  ll.lt.  not  .peclfled,  but  ..p              .   ,    .        ,          ...                         ,   .   .
                                        prond.d.  U.. ..i..ic .urr.y  p.rror.S!rH"Lfl       below the construction cost estimate. The closeness of the estimate

                                        sillin'.'iirnot'Vc'iuV. J" .""."ซ./"oteT'n       an<^ bid was ^ue *ฐ ^ne factฐrs discussed below.

                                        'i.ฃo..i ".r.""'^!™ .re™.iii" b.f".*ddre'"ed          The construction cost estimate was prepared in accordance with

                                        under ™th,r project, if ,t .ii.                  tne cOE's  guidance on estimating construction costs for  Super-
 Sabeurf.e. Or.ln.ee Control Durlnf Conetructlon      Sp.clflc.tlon. will require construction  to be        n    i      •       mi      *  i                    *    . >           f      *•
                                        .upended  durin.  period, of r.in  or  other       fund  projects.  The guidance suggests estimating costs from  the

                                        ci.in' """'rer^decont..^"'.^' ^"".V.iiw       contractor's perspective, including the  development of a critical
                                        .r... prior to r.lnf.ll erent.; If It doe. not           .     ^ijiji     j         ^>       *       ^jii
                                        do  .o,   then  nmoff  n.t  be  cou.cted.       path  method schedule and computing equipment and labor costs
                                        characterized  and  Banaged as  hazardous vaate.           • jซ     j i    i     i_ii  T^I •           i  •         ^.*              •
                                        luno. control win b. de.i^ruetauMcL"t..te^       lower level of protection could be justified. However,  it  was

                                        n.in, .  25 y..r .tor. .r.nt win b. u..d  'o       assumed (correctly, it turned out) that the contractor would be able
                                        d..lfn  runon  ud  runoff  ud  control..                 v         3 '              7
                                        te.por.ry truck .c.i.. ..y .1.0 b, in.,.ii.d m       JQ perform most of the work at Level C (air purifying respirators).
                                        thi. .r... W.ter uid power will be brought  lnr                                        \rjorf
                                        fro. off.it..                               A 15% factor was  added to equipment and labor costs to account

 ปt. ci..rini u>d D.bri. i—r.i                 so.,  bru.b ซd  tre..  no.  in contect with       for inefficiencies imposed by health and safety precautions.
                                        contuilneted  .oil.  cut  b.  re.ored  without                               r        J                     J *
                                        co.inj int.  cont.ct .ith tho.e .on.,  wh.re           of the $3.8 million estimated for construction, some $2.1 million
                                        possible, such an approach will be required.

                                        I^'cont'-ina^d "oTSl tii? JT .1.n.i.id"ซ       was ^or off-site transportation and disposal. Those who follow the

                                        hat.rdou. w..te ซ>d disposed offsite.               hazardous waste disposal business know that predictions regarding
 Site Closure                               lecord  of  DeciaIon  required backfilling  the          .      ,               .       .    .    -    _                    .     .
                                        site, covering with topaoii,  ซd  veget.ting the        disposal costs several  months in the future are uncertain at best.
                                        area. A on.lt* borrow area will ba Identified        _     .         .      -                            ,  ,             .
                                        on th. property con.iatซ>t vith the  wi.hea of        Based upon planned price increases reported  by waste industry
                                        the  present landowner.  Final cover  will be               .  .T    i  A  y     .*ซ,,,      ,         ^      i  ^
                                        designed to^inciud.^. .ini.u. o^ia  i™*ฃซ        financial analysts (specifically, based upon a Research Comment

                                        iimtrซiซ1.1' Thi.cor.py.?fviii b? covereTwltt        prepared by the firm of Merrill, Lynch, Pierce, Fenner &  Smith,
                                        a BinlBUB of 6 Inche. of topaoll, graded to 2X        f   f        J                ,.      .  .       .              n      .      ซ
                                        or it.  m .up., indigenou. v.get.tion will        Inc.), the  design team  adjusted its  estimated cost for disposal
                                        be  specified  consistent with  the  planting            '           ฐ             J         <ซ.ป..           .
                                        ซซซ•                                     upward by 22% over prices quoted by facilities at the time the cost

 L1.US of Excavation                         K.cord of Decision apeclfl.d excavation  within        CStimatC W3S prepared.
                                        -waata dlapoaal area- down  to bedrock,  and                       f   f
                                        removal  of  first  alx  Inches  of  bedrock.
                                        Hlnlalte amount of excavation beyond daalgnated


                                        to five  feet  in height,  and  Halt alopea above
                                        that height to 1H.1V.

 Subsurface Drainage Control                    Subdraln  Bedlia. nust  be free-draining (i.e.
                                        contain   no  fInea),  and   la  keyed  Into

                                        IX objective Is to cutoff groundwater flow Into
                                        waste disposal area.

 Project Access                              Hinlaiie  easeaentB on  other  properties.   Hew
                                        road  construction Bust be suitable  for heavy
                                        truck traffic (36  tona, axle  load of  16,000
                                        pounds),  but design  life  If only  6  months.
                                        Pave existing acceaa road (roughly 1,000 feet
                                        long).   New   acceas   roadway  to  be  denac
                                        aggregate base course.  MaxlBUB  grade for  new
                                        sections to be IX.  MinlBUB turning radlua  to
                                        be 50 feet.

 DecontaBlnatlon and Staging Area                 Locate at top of  rasp  leading froa excavation
                                        area. Provide iBpervlous surface sufficient  in
                                        size to  accoBBWdate a 400 square foot  staging

                                        Provide Beans of collecting and  holding  runoff
                                        froa Impervious area for 25 year, 24 hour a tons
                                        event.  Provide gradea of  between 1  and *
                                        percent,  and allow minimum  50 foot  outside
                                        turning radlua.

 Transportation                              Brldgea  along transportation  route  muat  be
                                        capable  of handling  3t ton  truck traffic.
                                        Dlapoaal truck, muat be weighed before entering
                                        ซnri after leaving contamination reduction tone.

 lite Closure                               Final gradea to Batch original contours where
                                        possible  and tie  In amoothly with adjoining
                                        areas.  KaxlBUM slope to  be  2H:  IV,  vlth a
                                        slope of 3D: IV preferred.  Hlnlmw slope to be
                                         IX  to promote  drainage.   ftackf111 aha11  be
                                         clean fill covered by topaoll, with no apeclal

                                         be ravegetated.
                                                                                                                         SITE REMEDIATION      299

-------
                                 Role of  Surface  Geophysics In
                 Developing Buried  Waste  Removal Specifications
                                                Norm N.  Hatch,  Jr.
                                                    CH2M HILL
                                                  Atlanta,  Georgia
                                                   William Owens
                                         Oak Ridge  National  Laboratory
                                               Oak Ridge, Tennessee
                                                  Samuel  Shannon
                                                   Patricia  Markey
                                                    CH2M HILL
                                                Gainesville,  Florida
ABSTRACT

  Surface geophysics was used to obtain information necessary to
develop technical specifications for removal of buried waste con-
tainers at two sites at Patrick Air Force Base (AFB), Florida, and
one site at Charleston AFB, South Carolina. One Patrick AFB site
reportedly was used in the early  1960s to bury pesticides in both
plastic and metal containers. The  other Patrick AFB site is located
in a section of a former base landfill where drums of waste oil are
known to be buried. The exact locations or numbers of buried
drums and containers at either site are unknown. The Charleston
AFB site was identified in May 1986 when Army Reserve person-
nel notified  base personnel that  they had located several drums
in a  wooded area south of the  base runway. The drums  were
believed  to have been unearthed in January 1986 during plowing
for pine seedling planting.
  A ground penetrating radar (GPR) survey and a magnetometer
survey were conducted at  the pesticide container burial site at
Patrick AFB; magnetometer surveys also were conducted at the
waste oil drum burial site at Patrick and at the Charleston  AFB
site in the area surrounding the exposed drums. Data from geo-
physical surveys are subject to interpretation, and anomalies identi-
fied  by the  surveys may be the  result of natural conditions or
objects. For this reason, hand excavations of major surface geo-
physic anomalies were conducted to identify the nature of the
anomaly targets  prior to development of technical specifications
for procurement of a cleanup contractor.
  At the pesticide container burial site on Patrick AFB, the hand
excavations  revealed that all of the anomalies were the result of
unexpected subsurface conditions (e.g., buried shell beds and tree
stumps, iron indurated rocks) rather  than buried drums or con-
tainers. In the case  of the Patrick AFB waste  oil drum site and
the Charleston AFB site, surface geophysics provided valuable in-
formation in delineating the boundaries of the areas containing
the buried drums.
  The experience at Patrick AFB demonstrates the limitations of
surface geophysics and the importance of target identification prior
to the procurement of a cleanup contractor to implement remedial
actions. In the case of the reported pesticide container burial site,
needless and costly excavations by a cleanup contractor  were
avoided.

INTRODUCTION
  The Oak Ridge National Laboratory  (ORNL) has interagency
agreements with the U.S. Air Force (USAF) to provide contracting
and technical assistance in coordinating hazardous waste projects

300   SITE REMEDIATION
at  USAF  installations.  Specifically,  the  agreements apply to
Installation Restoration Program (IRP) project work. ORNL, in
turn, has eight nationally recognized consultant engineering firms
under contract to assist in this work. CH2M HILL has a task order
contract with ORNL to provide IRP assistance to the USAF at
installations in the southeast.
  The IRP is intended to be consistent with CERCLA requirements
as augmented by SARA and implemented through the NCP. Many
of the SARA requirements specifically pertain to sites on the NPL.
Some requirements pertain to any facilities regardless of whether
the sites are included on the NPL.
  Potential contaminant sources are usually identified in a records
search, which is a preliminary assessment of contaminant presence
and possible migration. Further tasks may include a RI to evaluate
the magnitude and extent of contamination and a FS to screen and
evaluate remedial measures and to select the appropriate remedial
action that is cost-effective and adequately protects human health
and the environment. In some cases, immediate removal actions,
such as soil excavation or drum disposal, are identified during the
records search.
  Immediate removal actions, which can be identified at any time
during site remediation, sometimes require focused data collection
to define the scope of work and may require post-removal moni-
toring to assess the effectiveness of the removal. If no residual con-
tamination remains, the site may be removed from the  IRP. If
residual contamination remains, the potential for migration and
exposure is assessed to determine if remedial investigation activi-
ties are warranted. Immediate removal actions usually are under-
taken as a preventive action to remove a  well-defined potential
contaminant source before  extensive contaminant release and
migration  can occur.

IMMEDIATE REMOVAL ACTION APPROACH
  The basic approach to immediate removal actions involves the
preparation of technical specifications that base contracting per-
sonnel  can use to procure a qualified cleanup contract to imple-
ment the work. Close coordination with the base contracting office
during  this process is required to provide the technical specifications
in  the proper  format (each base contracting office  has different
requirements) for inclusion in the total procurement package. Front
end legal documents and general terms and conditions are specific
for each USAF installation, and this portion of the procurement
package is  assembled by the base contracting office.
  The following project-specific information must be provided to
the base contracting office:

-------
• A proposal schedule and bidder information requests
• Technical specifications
• Engineer's estimate of project cost and duration (for USAF
  internal use

  The proposal schedule includes an itemized bid list of lump sum
items and unit price items including estimated quantities for bidding
purposes. Bidder information requests may include: subcontractor
information, such as identifying the earthwork subcontractor;
hazardous waste transporter information including U.S. EPA and
state licenses;  hazardous waste disposal facility information
including U.S. EPA and state permits; laboratory certification if
laboratory analyses are included in the contract; a description of
previous experience on  similar work; and a description of the
approach proposed by the cleanup contractor for implementing
the immediate removal  action in accordance with the contract
requirements.
  The technical specifications are divided into two sections: general
requirements and specific requirements. The general requirements
specify the cleanup contractor's responsibilities regarding sequence
and schedule of operations, project meetings and coordination with
the base. General information about site conditions, temporary
utilities and facilities, site safety and security, site maintenance and
cleanup and submittals during construction is included.
  The specific requirements include site background information
and a  summary of waste characterization and  data collection
information,  a detailed  scope of  work, detailed requirements
pertaining to the execution of the work and the basis of payment
for the contract.
  In some cases, sufficient data may be available from previous
investigations to support the development of the technical specifi-
                           Figure 1
             Site DS-3, Patrick Air Force Base, Florida
cations. For both Patrick AFB and Charleston AFB, additional
data were required. In particular, surface geophysics played a major
and important role in the additional data collection activities.

FIELD INVESTIGATIONS AT PATRICK AFB
  Patrick AFB, encompassing approximately 1,800 acres, is located
near Cocoa Beach in Brevard County, Florida. According to the
Patrick AFB IRP Phase I report, excess or outdated pesticides had
been buried in a  200-ft diameter area designated as DS-3 (Fig 1).
The pesticide materials were believed to have been buried in both
plastic and metal containers in burial pits dug with a backhoe to
a depth of approximately 4 ft. The second site, LF-1, is located
in a section of a  former base landfill where waste oil drums were
known to be buried.
  GPR and/or magnetometer surface geophysical surveys were
conducted  at DS-3 and  LF-1 to  evaluate subsurface  conditions
without soil excavation. Magnetometers measure the local magnetic
field, and rapid changes in this field are indicative of the presence
of ferrous metal. GPR sends an energy impulse into the ground;
short, duration electromagnetic waves are reflected by materials
of contrasting dielectric properties back to the instrument.
  Changes in subsurface conditions detected by these  techniques
are called anomalies. Anomalies detected by magnetometry indi-
cate  the presence  of  buried ferrous metal objects. Anomalies
detected by GPR indicate changes in the depth or orientation of
subsurface reflecting horizons including soil layers, soil voids or
man-made objects. Buried debris such as tree trunks or construc-
tion wastes could produce GPR anomalies similar to a plastic con-
tainer.

Ground Penetrating Radar Survey
  A GPR survey was used at DS-3 to attempt to detect and map
buried materials or trenches. The survey was subcontracted to
TECHNOS, Inc. As  the radar antenna  was moved  along the
surface, the graphic  recorder produced a picture-like  display
showing a continuous profile along a traverse. Reflections occurred
from different soil horizons as well as from possible man-made
objects such as pipes and drums.
  The TECHNOS modular radar system is composed of a modi-
fied GSSI  4800  scientific system  and a GSSI Model SR8000H
graphic recorder. A monostatic (one transmitter/receiver element)
80 MHz antenna was selected for this site. The antenna was hand
towed across the  site and reference marks were put on the  records
every 20 ft.
  The depth range of the radar system was set at about 9 ft by
allowing a reflection travel time of 70 nanoseconds. A first approx-
imation model was used for depth calibration using typical values
of 5 nanoseconds/ft of travel time for the unsaturated material
and 10 nanoseconds/ft for the saturated zone. The water table was
approximately at a 3-ft deep, resulting in an effective penetration
of about 5.5 ft below the water table in the original  grade.
  A test of the GPR system was  conducted in the field to verify
the system's ability to locate or detect buried plastic  objects. A
test profile 20 ft long was established. The radar antenna was pulled
along this line and the radar profile produced showed  evenly dis-
tributed, essentially horizontal layers beneath the test strip. A hole
approximately 2  ft across and 3 ft deep was excavated and a 5-gal
plastic jug  of distilled water was buried half in and half out of
the water table. The radar antenna was again pulled over the test
traverse, yielding a radar response that indicated the presence of
a buried target or anomalous zone. To test the ability  of the unit
to locate disturbed soil, the 5-gal jug was excavated, the hole was
backfilled and the test strip was traversed a third time. The radar
response delineated the sides of the excavation. Thus, in this test,
the GPR system was able to locate buried plastic objects.

Magnetometer Survey
  A magnetometer was used at DS-3 and LF-1 to attempt to detect
and map the extent of buried ferrous metals (iron and steel). A
                                                                                                  SITE REMEDIATION     301

-------
vertical fluxgate gradiometer magnetometer was used to provide
a vertical gradient measurement of the magnetic  field. Magnetic
response is proportional to the mass of the ferrous targets and
inversely proportional to the  fourth power of the distance to a
discrete  target. The fluxgate  gradiometer magnetometer has a
maximum  sensitivity of  ฑ7.6 gammas/ft  at  a  full scale. The
maximum resolution is approximately ฑ0.1 gammas/ft, according
to TECHNOS.
  The magnetometer was carried with the bottom sensor at a height
of approximately 2 ft above the ground surface to minimize the
effect of small surface targets. Parallel traverse lines were run south
to north at a  10-foot line spacing at both DS-3 and LF-1. Data
were continuously recorded on  an analog strip-chart recorder along
each traverse. A reference mark was made on the strip-chart record
every 20 ft for location  purposes. The  process of making station
marks at selected intervals eliminates cumulative position errors
that otherwise may occur in positioning along the magnetometer
traverse.

Conclusions of Geophysical Survey
  The GPR survey identified both localized anomalies and possible
burial trenches. The localized anomalies  occurred at random at
DS-3 in three areas, thought  to be possible burial trenches: a
northernmost anomaly, a central anomaly, and a  south, western
anomaly (see Fig 2). The magnetometer survey at DS-3 identified
only two anomalous areas. All other areas had readings that were
typical background values with little or  no response. The magnetic
anomalies were believed to represent buried drums or scrap metal
that are often associated with disposal sites.
-if
        LEGEND

        Radar Traverse Locations

        Radar Anomaly—Specific Target

        Radar Anomaly-PoeslMa Trench Areas
       • Toe of Slope
        (Baaed on field observations)
         Radar Anomaly-Cause Unknown
                                              Soil* In !ซ•!
                                                                     The  northernmost  anomaly was estimated  to  be 3,300 ft3
                                                                   (19,800 ft3). The radar data indicated a possible burial trench at
                                                                   this site. Furthermore, the site was thought to possibly contain me-
                                                                   tal drums because of the concurrent magnetic anomaly, represent-
                                                                   ing a surface area of approximately 1,300 ft2. The spatial extent
                                                                   of the anomaly was based upon the radar data.
                                                                     The central anomaly was estimated to be 3,500 ft2 (21,000 ft1).
                                                                   The radar data also indicated a possible burial trench at this site.
                                                                   Furthermore, the western portion of the site was  thought to possi-
                                                                   bly contain metal drums,  as indicated by the magnetic anomaly
                                                                   with a  surface area of 600 ft2.
                                                                     The southwestern anomaly was estimated to be 830 ft2 (5,000
                                                                   ft3). The radar data indicated disturbed soil or a possible trench-
                                                                   like feature here. There was no magnetic anomaly at  this site.
                                                                     Depths were estimated from the depth of the maximum radar
                                                                   anomaly and/or the magnetic data. Both resulted in a depth of
                                                                               LEGEND
                                                                                       Magnetometer Traverse Locations—See Note 1
                                                                                       Magnetic Anomaly—See Note 2
                            Figure 2
    Site DS-3 Radar Anomaly Map, Patrick Air Force Base, Florida
                                                                             Noln  1 Buried drums are located In the
                                                                                    magnetic anomaly areas The Contractor
                                                                                    shall locate and remove the drums
                                                                                    The Contractor shall alto remove
                                                                                    contaminated soil as directed by the
                                                                                    Contracting OHlcer

                                                                                   2 The magnetometer traverse lines are not
                                                                                     based on true north coordinates They are
                                                                                     developed from point A near the comer
                                                                                     ol building 111?


                                                                                              Figure 3
                                                                     Site LF-1 Magnetic Anomaly Map, Patrick Air Force Base, Florida
302     SITE REMEDIATION

-------
approximately 6 ft as typical. The volume was calculated as the
area times the depth, which is a conservative estimate for volume.
  In addition to the three large anomalies, numerous localized
anomalies were identified by radar. No magnetic anomalies were
associated with these localized, random anomalies, however, and
they may occur from natural conditions.
  All of the anomalies identified as possible burial locations  at
this site (including supposed trenches, localized radar targets, and
magnetic anomalies) occupy a net surface area of approximately
7,600 ft2. Using the depth data given above, about 1,700 yd3  of
material that might have required excavation to locate any buried
waste containers at DS-3. Because the water table beneath the site
is as shallow as approximately 2.5 ft, all of the material of concern
would probably be wet.  This  would make excavation/removal
activities more difficult.
  The magnetometer survey at LF-1, where waste oil drums were
known to be present, identified one large anomaly and one smaller
anomaly (see Fig 3). All other areas on the site had readings that
were typical of background values with  little or no response.
  TirgMI

0,-U I
  .'.'•'
-H"
     LEGEND

    - Radar Traverse Locations

 O   Radar Anomaly-Specific Target

>?Tv!;i: Radar Anomaly—Possible Trench Areas


     Toe ol Slope
     (Based on field observations)

     Radar Anomaly—Cause Unknown


      Excavated Area
                                       tc-JE:
                                            Seal* In I.el
                                                                    The major anomaly represented a surface area of approximately
                                                                  4,800 ft2. The entire anomaly had magnetic values that indicated
                                                                  metal was buried beneath the surface at depths ranging from a few
                                                                  inches to as deep as 10 ft. If this anomaly were the result of buried
                                                                  drums, it is likely that about 10 drums would be found, assuming
                                                                  a depth of 7 ft. The anomaly may also have been caused by other
                                                                  scrap metal buried beneath the surface.

                                                                  Results of Excavations at DS-3
                                                                    Four of the anomalies detected by surface geophysical surveys
                                                                  at the 200-ft diameter DS-3 site were excavated to evaluate the cause
                                                                  of each anomaly. Excavations were conducted by hand shoveling
                                                                  down  to the  water table (see Fig 4).
                                                                    Target 1 was selected for excavation because it was detected as
                                                                  a group of four discrete GPR anomalies. Target  2 was detected
                                                                  as a magnetic anomaly and two discrete GPR anomalies. Target
                                                                  3 was  detected as both a magnetic and GPR anomaly. Target 4
                                                                  was detected as two discrete GPR anomalies.
                                                                    At Target 1, sand and shell dredge spoil was encountered in the
                                                                  excavation. Thin clayey zones occurred above the water table, and
                                                                  a coarse shell layer occurred at an approximate depth of 2 to 2.5
                                                                  ft. Water was encountered about 2 ft below the ground surface.
                                                                  At Target  2,  about 6 in of loose sand was found to overlay 8 to
                                                                  10 in  of hard asphaltic sand. A corrugated metal culvert was
                                                                  encountered  under an old road bed at 4 to 4.5 ft in  depth. At
                                                                  Targets 3 and 4, wood fragments (roots and trunks)  were unearthed
                                                                  just above the water table at an approximate depth of 2 ft. Probing
                                                                  below the excavation bottom revealed the presence of a hard layer
                                                                  which in Target 3 was a shell bed and, in Target 4, was a dense
                                                                  clay bed. Iron-rich concretions and iron staining of the soil were
                                                                  also observed in  both  excavations.
                                                                    Continuous monitoring in each excavation with an HNu photo-
                                                                  ionizing organic vapor detector did not indicate the presence of
                                                                  contamination at  the excavated areas. The holes were all backfilled
                                                                  to original grade.
                                                                    No  pesticide container was  found during the excavations. No
                                                                  discolored or stained soils which could have been attributed to
                                                                  pesticide compounds were found.

                                                                  FIELD INVESTIGATION AT CHARLESTON AFB,
                                                                  NORTH FIELD DRUM SITE
                                                                    The North Field drum site  was defined as a rectangular area
                                                                  measuring 200 ft  by 400 ft, centered on the exposed 55-gal drums
                                                                  of waste (Fig 5). What appeared to be a possible burial trench was
                                                                  located about 100 ft east of the exposed drums. This feature was
                                                                  included in the study area, which was surveyed and marked into
                                                                  a grid for data collection.
                                                                           Grid Emuntion
                                                                                                                   To Runwayi


                                                                                                                   o/ 
-------
Procedures
  A magnetometer survey was performed to locate buried ferrous
metal objects within the study area. Field procedures consisted of
data  collection, primary analysis and additional data collection.
Magnetometer data were initially collected  on Dec. 17,  1986.
Additional measurements were taken along the extreme north and
south ends of each line to be used to correct for diurnal drift if
necessary. Field notes included the time at the beginning and end
of each line, the total  field magnetic reading and comments con-
cerning the location of observed scrap metal, drums, roads and
other features.

Magnetometer Data Reduction and Interpretation
  The  earth's field was  estimated to  be slightly above 52,000
gammas at the North Field drum site. A constant of 52,000 gammas
was subtracted from the data so that the residual (anomalous) Held
was mainly positive. Inspection of the uncorrected data indicated
that any diurnal drift that may have been present was small com-
pared with the observed anomalies and no attempt was made to
correct for the diurnal drift.
  The  absence of magnetic rock types at the site and  the strength
of the observed anomalies indicated that the anomalies were caused
by buried man-made metals. It was impossible to predict by analysis
of the magnetometer data alone whether the anomalies were caused
by drums or other scrap metal. Estimation of the volume of metal
detected was not possible either, because of the various factors that
affect  the anomaly amplitude.
  The  data were  entered into a computer the same day they were
collected and hand contoured to determine if any anomalies found
at the  edge of the grid needed additional measurements to fully
define their extent. Two such areas were found and additional work
was performed the next morning.
  The  data were presented as a contour map that indicated  the
spatial relationship of the anomalies and to some  extent allowed
an  interpretation of the source positions. The source areas were
determined from analysis of both the contour maps and profiled
data plots. Within each interpreted source area, the buried  metal
may have  been unevenly distributed.  However,  the resolution
permitted by a 10-ft grid of data prevented further breakdown of
the source areas.
  Seventeen anomalous areas that were thought to represent areas
of buried metal were delineated through analysis of the magneto-
meter data. Of the anomalies identified, seven were considered to
be priority targets for further investigation, as they had magnetic
anomalies  greater than 100 gammas that could not be explained
by  metal seen on the surface (Fig 6). Based on interpretation of
magnetometer data, depths to the sources of all the anomalies were
believed to be less than 10 ft. The remaining anomalies exhibited
minimal magnetic deflections and were not considered of further
interest.
 Anomaly Investigations
  The  seven anomalies chosen  for further investigation were
manually excavated in an attempt to identify  the source of the
magnetic field deflection. Manual excavation was selected to reduce
the chance that intact waste containers, if present, would be broken
or punctured and release pollutants into the environment. Anoma-
lies farthest from the known drum burial site were excavated first
to allow refinement of excavation and monitoring techniques during
less hazardous conditions.
  Excavation methodology consisted of shoveling in Safety Level
D until either  waste  was  encountered  and identified  as non-
hazardous or until potentially hazardous conditions were identi-
fied. HNu photoionization air monitoring for organic vapors was
conducted during  the  level D work.
  During potentially hazardous conditions, including attempts to
open and sample contents of unearthed 55-gaJ drums and hand
augering and collecting soil samples from beneath leaking 55-gai
drums, work was conducted in Safety Level B. MSA positive pres-
sure, dual-purpose, self-contained breathing apparatus equipped
with 150-ft long air lines were used. The only conditions encoun-
tered requiring Safety  Level B were at the central anomaly at the
known drum disposal site. All other work was conducted in Safety
                              Table 1
                Descriptions of Anomaly Excavations
                    At (fee North Field Drum Site
           Excavation
            Location
            200 e
            1H H
      pit conEaining b**r can*, bullat caainoe, ration
cana bar 1*4 to • depth of 5 ft.  Top :2 ft of Mil it
•ottled red. yellow, and brwn Mndy clay alied with
clayey ***<*- * vwry denaa. hard-parked layer of th* MM
•oil •Bt*ndt trtm 2 to ) ft and <*at lavediataJr laaderleia
by 9arb**e-  frelov i*~ tarbao*. *t 5 ft. waa aolBt. dark
rwSdtah brown a*ndy clay.

•*tw*ซn four ซnd **v*ซ wซlded ateel tn^le iron rack* or
•tand* roughly 1 ft aquar* by * ft Ion? <**r* encotiattared
at a depth of between 7 to 4 ft below tปe surface. Soil
la f ia*™o;rain*d tafl. loo** to enderately dona*, clayey
•and. Tfปa root* up to 2 In. in diaai. **rป growing
throoah and in*to* IN* Mtel la^> and br*c*ซ of to* rack*.
ป V-to fr-ft lonq piece of telephone cable wai also
localad.

A 1-iach die*, wire rope ca&l* about 2) ft lono. Betw*a
12 and IS ft of th* cable were bur lad within fc iochet of
the surface. Several holaa w*re dug to a depth of about 2
ft. and all ซ**r* hot to* ad in orano* clayey aand. * avtai
praba Maa uaซd to aound thซ ancir* aaoaMly on I to 1 ft
cซntarป, and no burtad hard ob)ปcta wซra dvtactari. Soil
waa looซa to Moderately dans* tan And brtMB clayey ปwd
ovvrtylng oranqe clayey a*nd.

Garbage pit cvmtalnin^ ateel o> iroa, beer and aoda caซs
without pop-top*, a aoda bottle dated lปM, and other
•atal and broken alas* wast*.  Soil ซaป loeae branป to
crania sandy clay to an IB-inch depth, underlain by a
den** darker layer of *andy clay contatmrvq root*. K
den*e on* 1-ft thick layer of oranqiah aandy clay overlay*
a 2-ft thick depotit of can* and bottlei.

Carbao* pit containing can*, bottlea, and rubb*r~clad
wire, Carba9e owflain by about 3 ft of v*ry denct broim
to orarwj* clayey ป*nd.  Ttie toll wซป !•ป• 4*n>* around the
wast* and v*ry denaely packed ov*rlyinq the ^arttaoe.

Two cruihed 30-oal ซetal qarbaoe ce.nt with lldi burled
batw**n !• and SO Inch** d*ซp.  Soil waa tan to brown
clayey *and BlKCd **lth dark reddish brown randy clay.
Hater encountered at abtn.t < ft below land lurface.

•eautanta of decayed Metal-bound wooden bomet )uซt above
the water table at 3-S ft below land aurface. Soil waป
brown and tan Moderately denae clayey aand.
                             Figure 6
         Priority Magnetic Anomalies, North Field Drum Site
                                                                                     10} t
                                                                                      ซ N
                                                                                     140 C
                                                                                     m H
                                                                                     111 K
                                                                                      JOO M
            100 K     Tvn S*-9*l druMB.  rivป drwt ซMrซ un*ซrth*d thซc Mr*
            Hi B     lซylnซ on th.lt .idn Tป.t th. pnซloviil> upoปd dnปซ
                     Tn
                     1 ft dซซp.  Soil w.i 100*9 clky.y tund.


encountered at )


* ft. Soil waa tan to brown clayey und.
304    SITE REMEDIATION

-------
Level D.
  Table 1 presents the findings of the excavations performed at
the seven magnetic anomalies classified as priority targets requiring
further investigation. Buried metal wastes were found at six of the
anomalies excavated. Anomaly D appeared to be related solely to
the surface metal, a wire rope cable, which was much longer and
contained more metal than was originally thought. Wastes thought
to be potentially hazardous were located at anomalies K and M
(Table 1). The remaining four anomalies appeared to be caused
by buried metal wastes such as scrap metal or garbage. The only
anomaly sources  located that  were believed to  pose a threat or
potential threat to the environment were the wastes at anomalies
K and M. Anomaly was the only location where buried drums were
found (Fig 7).
 NOTES Dfunwt.J. 3.1. ind fi irซ MPOMTJ
     Drums ซ. 5. B. 7. ind 10 tn boriad
     on lhปlf tidM ind appMr 10 D* undamlgซ41.

     Point A to loeitod II mtgrwlointutr grid
     loullon Eul 200, North 100. Location! ol
     other drumt ini iperoxlrrat* only and ซrป
     nottgiul*
                           Figure 7
 Approximate Location of Drums at Anomaly M, North Field Drum Site
CONCLUSION
  Surface geophysics provided valuable information on the poten-
tial locations and extent of buried containers and debris. Truthing
of the anomalies and targets identified by the geophysical surveys
was essential to fill information gaps and to develop a meaningful
set of technical specifications for removal of buried waste materials.
Hand digs at selected major anomalies verified the presence and
exact locations of buried drums and other buried waste at the
Charleston AFB North Field drum  site.  This information will
minimize the expensive "exploratory" excavation time and effort
in the field  by the cleanup contractor,  which could  have  been
extensive.
  For the Patrick AFB reported pesticide container burial site, the
data  collection program  of surface  geophysics,  truthing  of
anomalies and targets and soil and groundwater sampling con-
cluded that no buried pesticide containers were present in the sus-
pect area or vicinity. A decision document for the site summarized
all available information and the rationale leading to the decision
that an immediate removal action was not appropriate.
  Truthing of the geophysical survey anomalies in this case saved
time, effort and cost. If truthing had not been done, technical
specifications  would have been developed based on the cleanup
contractor excavating all of the major anomalies and targets and
expecting to find buried pesticide containers. This excavation would
likely have been performed using Level B health and safety pre-
cautions (self, contained breathing  apparatus)  because of the
unknown nature and potential hazards of the container contents.
The excavation could have encompassed approximately 1,700 yd3
of soil at a unit price of about $50/yd3. Additional expenses
would have been incurred with the development of the technical
specifications and cleanup contractor mobilization, demobilization
and other lump sum costs. Total project costs would  have  been
in the range of $100,000 to $150,000, and the result would  have
been the discovery that there were no buried pesticide  containers
at the site.
   The surface geophysical data, taken alone, indicated the presence
of buried material, possibly drums. The truthing of the geophysi-
cal anomalies provided the essential information that led to the
decision not to proceed with the immediate removal action.
                                                                                                   SITE REMEDIATION    305

-------
                                 Immediate  Removal Activities
                   at  a  Dioxin-Contaminated  Mobile Home Park

                                                 Russell B. Krohn
                                            Tetra Tech, Incorporated
                                                Kansas City, Kansas
                                               James R. MacDonald
                                    U. S.  Environmental Protection Agency
                            Region VII, Emergency Planning and Response Branch
                                               Kansas City, Kansas
ABSTRACT
  Quail Run Mobile Manor is a mobile home park located near
Gray Summit, Missouri. In the early 1970s, the area was contami-
nated with 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) when the
main gravel road was sprayed with TCDD-contaminated waste oil
as a dust control measure. This common practice of the time created
many of the dioxin sites in Missouri. The Quail Run site was first
sampled in early 1983 and temporary relocation was offered to resi-
dents based on the levels of dioxin present in the soil (up to 2,200
ug/kg). Subsequent investigation revealed that contamination had
spread from the road throughout the site via tracking, wind and
water erosion and the use of contaminated soil as fill at nearby
locations. Affected areas included 28 mobile homes, a house, the
roof of an underground home, substrate under an equipment shed,
1,400 ft of highway shoulder and approximately 5 acres of the park
property.
  In response to the threat to human health and the environment,
the U.S. EPA initiated a CERCLA. Immediate Removal Action
in February 1985. This mitigative action was designed with the
cooperation of state  and  federal health agencies. It addressed
dwelling and debris decontamination as well as excavation and con-
tainerization of contaminated soils for on-site storage and future
disposal. Many of the decontamination, excavation and storage
methods utilized were not new. Innovative techniques for cleanup
of soil and structures  were continuously implemented to increase
the efficiency and effectiveness of the removal project. Personnel
safety and potential exposure to the public were major concerns
and were  addressed by an extensive air monitoring program.
  Structures and soil were extensively sampled to define areas of
contamination prior to cleanup activities. Sampling was performed
throughout the duration of the project to confirm the effective-
ness of removal efforts.  Soil  sampling incorporated methods that
statistically evaluated dioxin concentrations to the upper 95% con-
fidence level.
  Approximately 600 yd3 of concrete and debris were decontami-
nated; 16,000 yd3 of soil were excavated; and all property was re-
stored  to  its  original  condition.  Public  relations and
intergovernmental communications were critical to alleviate public
concern and to generate support  for the project  throughout this
highly publicized 2-year removal action. The methods and poli-
cies established during the Quail Run removal served as models
for later actions of similar nature.

INTRODUCTION
  Quail Run Mobile Manor is a mobile home  park located in
Franklin County, Missouri near Gray Summit. The road through
the park was sprayed for dust control with dioxin-contaminated
waste oil in the early 1970s. Soil from the site was first sampled
in February 1983 and numerous additional sampling efforts were
conducted  prior to the initiation of full-scale removal activities.
Contaminated paniculate spread through tracking, wind and water
erosion, greatly increasing the areal extent of contamination. Soil
from the park was used as fill dirt, thereby increasing the spread
of contamination. Affected areas included the park, the Chlanda
property, the Mahaney earth home, road shoulders and adjoining
properties  of Highway  100 (Figure 1).
  UStDCKnAL HCMU
                         Figure 1
                  Quail Run Mobile Manor
              Gray Summit, Missouri (Not to Scale)

  In response to the threat to public health, the CDC issued a
Health Advisory in May 1983, recommending relocation of resi-
dents of Quail Run Mobile Manor. Twenty-one families immedi-
ately accepted temporary relocation benefits through an Interagency
Agreement (IAG) between FEMA and U.S. EPA. Following that
relocation, U.S. EPA conducted Phase I mitigative actions which
included fencing the park and paving the entrance to encapsulate
contamination. Twenty-four hour security was placed on the site
to prevent unauthorized entry.
  In January 1984, Quail Run was included with five other eastern
Missouri dioxin sites in a ROD authorizing planning for an Interim
Central Storage Facility to be located at  Times Beach. This plan
was later abandoned due to public and political opposition. At that
time, a Mitigation Plan was developed specifically for Quail Run.
The plan proposed consolidation of contaminated materials with
temporary storage in buildings that had impervious pads and were
located in a secure area on-site. The plan also outlined decontami-
nation procedures to be used on mobile homes located on-site. The
306    SITE REMEDIATION

-------
Action Memorandum justifying this Immediate Removal under the
provisions of  CERCLA was signed  by the Acting Assistant
Administrator  in February 1985.

IMMEDIATE  REMOVAL ACTIVITIES
Health and Safety
  The U.S. EPA OSC had the overall responsibility for implemen-
tation of the site health and safety plans. The field teams operated
in Level C protection. A contamination reduction zone with decon-
tamination stations for personnel and equipment was established
between the contaminated zone and the clean zone. A decontami-
nation trailer was available for showers before lunch and prior to
departure at the end of the day.  Prior to assignment at the Quail
Run site, all personnel underwent a minimum  of  40 hours  of
training in Hazardous Materials Evaluation and  Response and a
complete physical examination for baseline monitoring.  A nurse
was on-site for worker weight measurements, blood pressure and
oral temperature approximately four times  each day during the
hotter summer months. Special  emphasis was placed on preven-
tion of heat stress through training, monitoring of liquid intake,
records  of vital signs and control of break periods. This plan was
implemented because of high temperatures, humidity and required
protective clothing, all of which contribute to physical stress of
workers.

Mobile  Homes
  Twenty-eight mobile homes were to be subjected to decontami-
nation and restoration actions, after which they were to be returned
to the owner or stored  until  the Quail  Run park excavation and
subsequent restoration was completed. The goal of the trailer resto-
ration was to return the individual units to their original  condi-
tion prior to the discovery of dioxin and relocation in May 1983.
  In  preparation for anticipated removal activities, U.S. EPA
representatives met with mobile home owners beginning in Decem-
ber 1984 to discuss proposed trailer decontamination and specifics
concerning their personal property located at Quail Run. Once the
Immediate Removal action was authorized, on-site video-recorded
inspections of trailers with the  owners were conducted by U.S.
EPA, U.S. EPA contractors and a mobile home restoration con-
tractor. The mobile home contractor was to restore the homes,
replacing interior furnishings and set up the trailers for occupancy.
  Once a homeowner participated in a video inspection to record
damages from abandonment, furniture to be replaced  and any
special  considerations  to  be taken,  an access  agreement was
developed for the owner to grant permission to perform the decon-
tamination and restoration. The first agreements were signed in
February 1985, though in spite of continual efforts by state and
federal representatives,  many of the residents did not grant access
until the summer of 1985. Access to the last unit was granted in
September 1985. The lack of simultaneous access to all trailers sig-
nificantly decreased efficiency and increased costs to decontaminate
the homes. At the time Immediate Removal activities were initiated,
five mobile homes were occupied. Relocation of all  residents of
Quail Run was not  accomplished until July 1985, hampering
efficient work  efforts.
  Procedures  for decontaminating  the  mobile homes  were
developed by the U.S. EPA and were reviewed by the U.S. CDC
and the Missouri Department of Health (DOH). Once a mobile
home was decontaminated and sampled, the Missouri DOH issued
a certification letter for each unit, stating that the unit had been
cleaned  and sampled to their specifications (with no detectable
dioxin) and did not represent  a significant threat to human health.
The Missouri DOH  letter to  each homeowner was accompanied
with a memorandum, describing all samples collected in the home
and a copy of the analytical results.
  Decontamination procedures  began with stripping the interior
of the home of all porous fabric material (e.g. carpeting, drapes,
furniture). This material was disposed of as contaminated. The
floorboard insulation and ductwork were removed. Non-porous
material such as the ducts, skirting, bedsprings, etc., was routinely
washed and disposed of after being sampled and validated clean.
  Once the interior was stripped, final decontamination included
a series of vacuum, wash and wipe down cycles. A visual inspec-
tion was conducted by a consortium of management personnel to
determine whether the unit had been  sufficiently cleaned for
sampling. Once the trailer was approved by the inspection team,
extensive wipe sampling on the interior and exterior was performed
to analytically verify the decontamination efficiency. Sampling
protocol  was based on a procedure outlined by the U.S. CDC,
which called for horizontal and vertical wipe samples from each
room plus one sample from a food preparation surface (counter-
top) in the kitchen.
  Cleaning was done on a trial basis in March 1985 in the mobile
home that had  the highest documented dioxin contamination level
(11.51 ug/kg; vacuum sample). This home was extensively sampled
before and after cleaning,  providing data that showed the units
could be effectively decontaminated. Following the-successful trial
decontamination, cleaning of the 27 additional mobile homes began
and was completed 4 months later, with the exception of one unit
for which access was not granted until later.
  Although not required by the CDC, four external wipe samples
were obtained from each unit to satisfy city officials that external
contamination was not present on the units being restored in their
municipality. All data from interior and exterior samples was vali-
dated prior to  off-site transport  of each unit. A package of data
containing the sample description memorandum, trailer sketch and
laboratory data transmittal was sent to the city prior to departure
of the unit for the restoration site. All finals samples had to show
no TCDD present at a minimum detection level of 4 pg/cm2. Of
approximately  375 wipe samples taken from the 28 mobile homes,
only two samples showed detectable levels of dioxin after initial
cleaning. Both of these samples  were collected from the rusted
exterior I-beams, which were more difficult to thoroughly clean.
The highest value found was 0.68 pg/cm2.
  A major issue concerning decontamination of the trailers was
whether wall insulation represented a reservoir for contamination.
Preliminary sampling of insulation from several trailers, including
the demonstration unit, indicated that significant contamination
was not present, although the insulation and wall spaces generally
appeared to be dirty. Origin of this dirt could have been from
manufacturing and transport to the site or subsequent degrada-
tion of the insulation. Approximately 120 insulation samples from
the trailers was collected. Insulation in eight walls of six trailers
were removed  as a result of detected contamination, although
insulation in only one wall exceeded 1 ug/kg. This wall was located
next to a pothole in  the road that passing  vehicles continually
splashed  with  contaminated mud,  which may account  for  the
unusually high concentration of dioxin in this insulation.

Rental House Decontamination
  The Quail Run site had an old farmhouse near the entrance that
had been used as a rental property. Decontamination efforts used
in this  structure included processes similar to those used on the
mobile homes. The interior was stripped  of porous materials
(including attic insulation; cabinets and other structures) to allow
thorough cleaning. All surfaces were vacuumed and washed.
Decontamination was begun in Fall 1985 and continued periodi-
cally until completion in Spring 1986. Once cleaned, the structure
was sampled inside and out to certify it clean. Once clean, a local
contractor restored the structure for future habitation.

Storage Facilities and Access Road
  The Mitigation Plan proposed containerization of excavated con-
taminated soil  in woven polypropylene bags with a polyethylene
liner. These bags were to be stored in a 5,000-ft2 totally enclosed,
steel-sided soil storage structures with impermeable, asphaltic con-
crete floors. The foundations of the structures included a 6-in. berm
around the floor perimeter. The floor sloped inward to a poured
                                                                                                SITE REMEDIATION    307

-------
concrete catch basin with grates for collection of any potentially
contaminated liquids leaking from the bags.
  The original estimates of soil to be excavated indicated that a
total of seven to 10 structures would be needed. However, due to
greater area! and vertical extent of contamination than expected,
16 buildings were constructed  and filled by project completion.
Bags were stacked in a pyramidal formation to prevent falling; this
configuration required a great  deal  of extra space, particularly
because the bags varied in loaded shape and amount of fill. A total
of 12,222  two yd3 bags of soil and some trailer debris were stored
in the structures.
  An access road was proposed and built to the storage facility
area. The  entrance from Highway 100 was located on the west side
of the adjoining Chlanda property, crossing into the Quail  Run
property and up the back hill where the  storage facility complex
is located. Although the proposed road  location was planned to
be located in non-contaminated areas, pre-construction sampling
proved the need for excavation  of 50,000  ft2 of contaminated  area
within Quail Run prior to road construction. Road construction
proceeded in non-contaminated areas during excavation. Access
road plans included a transfer pad and access spur for movement
of bags of material from the hot zone in Quail Run to the clean
area.  Unusually wet weather throughout the spring and summer
continually delayed construction of the road and storage facilities.
  Construction of the basic storage  facility complex grades was
completed in July 1986 and the first structure was completed in
August. Additional pads and  buildings  were built continuously
thereafter. Rain continued to hamper efforts, particularly construc-
tion which required drier conditions  for a good base. If the pads
were rocked and  paved while the underlying base was wet,  soft
spots resulted and were subject to cracking under heavy equipment.

Excavation Procedures
  Sampling to define  areas of contamination of greater than
1 ug/kg at the 95%  upper confidence level (UCL)1 in the Quail
Run and Chlanda properties was initiated in April 1985.  Sample
grid maps were developed, using property boundaries and natural
features (such as the old Quail Run road) as bases for the grid
arrangements.  Whenever possible, grids of areas 5,000 ft2 were
incorporated as directed by the sampling plan. Beginning in  July
1985, a GC/MS/MS unit was used for soil analyses. Costs for soil
samples were significantly reduced as were turnaround times2. It
was at  this  time that definitive sampling of the Highway 100
shoulders and the remainder  of Quail Run  was aggressively
undertaken and essentially completed in August. Approximately
800 soil samples were collected and analyzed from pre- and post-
excavation areas  for the duration of the project.
  Excavation of soil  was performed by either a trackhoe or back-
hoe with a smooth edged bucket. A cut of pre-determined depth
was made and  soil was deposited in a custom fabricated hopper
that funneled the material into flexible, woven, polypropylene bags
2 yd3 in volume.  These bags were rated  at 6,000 Ib capacity and
had an  8-mil plastic liner.  Both the inner and outer containment
bags were sealed  after filling and a crane removed the bag from
the hopper and placed them on a dump truck. The bags were then
transported to the storage facilities where they were unloaded and
placed in storage.
   Following excavation in the vicinity of the access road, excava-
tion was  begun  in July  1985  on  the  Chlanda property.  This
property,  which adjoins Quail Run (Chlanda formerly owned Quail
Run), became contaminated when the owner regraded the park road
and used the material for fill underneath the equipment shed. Con-
tamination spread during the fill operation and through subsequent
vehicular tracking and erosion. The Chlanda equipment shed was
dismantled after decontamination of the entire structure was com-
pleted and wood shavings of the internal framework proved the
material clean  for reuse. Because of the logistical problems, the
building was relocated on a new pad approximately 300 ft from
its original location.  The original pad was removed and underly-
ing and adjacent soils were removed. 686 bags were filled with con-
taminated soils.
  Prior to the discovery of dioxin at the site the Mahaney under-
ground home had received contaminated soil from Quail Run to
cover the roof. This remedial operation required complete removal
of approximately 24 in. of fill from the roof as well as the insulating
and waterproofing materials from the entire outside of the con-
crete shell. Although the possibility of damage to the structure
existed during the removal, extreme precautions were taken and
no problems were encountered. A total of 556 bags of material
were removed (722 yd3). The local firm that had built the home
was subcontracted to restore the waterproofing and insulation after
the removal was completed.
  Following the completion of the Mahaney project, excavation
resumed within the Quail Run property. A second excavation crew
was added in September,  greatly expediting progress. Both crews
continued with this effort until November, when all excavation
activities ceased due to continuing rain. Excavation did not resume
until 1986 activities were initiated in April 1986 and continued until
July. A total  of  approximately  9,930 2-yd3  bags of soil  were
removed  from approximately 5 acres in the park.
  Although rainy weather was common during the 1985 activities,
periods of dry weather necessitated implementation of dust con-
trol measures. These measures generally involved use of garden
or use of a 2-in. firehose to apply water to the excavation  area.
Occasionally, sprinklers were used overnight in areas to be  exca-
vated the following day. The old Quail Run road, which was highly
contaminated underneath the surface, was heavily used for trans-
port of bags from the north end of the park to the transfer pad.
This usage broke up the surface and with dry weather represented
a major potential source of airborne contaminated particulate. A
water tanker trailer was used to spray the road surface to prevent
dusty conditions.
   A major task associated with the excavation of the park was the
removal,  pressure washing and sampling of chunks of concrete that
were formerly patios and trailer runners within contaminated areas
of the  park. Approximately  600 yd3 of concrete required this
treatment. Several trials were conducted to determine what method
of cleaning and sampling was most efficient. Suspension of the
concrete  with a wire noose from a crane while being pressure
washed with a detergent solution proved to be effective, although
slow and tedious.  Once a pile of approximately 15-20 yd3 of con-
crete was cleaned, it would be sampled by chipping approximately
50 surface pieces  from throughout the pile, if analysis indicated
the pile was clean (< ug/kg TCDD), it was landfilled on-site.  From
mid-November to shutdown in February, concrete decontamina-
tion was  the only  significant site activity. Concrete decontamina-
tion was  begun again in April  1986 when full-scale site activities
resumed  after the 2-month winter break. Much of the later con-
crete was placed in a local sanitary landfill after being sampled
and designated clean. This  effort saved storage and future  treat-
ment/disposal costs.
   Approximately 700 ft  of each shoulder of Highway 100 were
contaminated with dioxin at levels greater than 1 ug/kg. This con-
tamination was spread primarily by vehicles tracking contaminated
soil from the park, although erosion likely spread contamination
further. The contaminated sections were each 15 ft wide, with  a
total contamination area 21,000 ft2 of shoulder to be removed.
Excavation of the shoulders  began  on the south  shoulder in
November 1985; however, rain forced the shutdown of this  phase
of remediation. Excavation  of the shoulders could not resume until
Spring 1986 and was completed in mid-May. A total of 1,040 bags
(1,352 yd3) of contaminated soil were removed from the highway
shoulders. Restoration began once excavation was completed.
   The shoulder section at the entrance of the park required exca-
vation to  4-ft depths, which  necessitated  the installation of  a
temporary bypass of the water main in the section. Sampling
attempts to identify a suspected isolated area of high-level con-
tamination in the section and thus avoid relocation of the line  failed.
 308    SITE REMEDIATION

-------
This section was likely contaminated to extreme depths by the
installation of subsurface utilities through the entrance shoulder.
  Concerns were high for the safety of the workers as well as the
travelling  public  on  Highway 100 during  excavation  of the
shoulders. An advance warning sign setup was used at all times,
in accordance with Missouri State Highway Department work
permits. While excavation was actually occurring, one lane of the
highway was shut down to traffic so equipment could be staged
and bags could safely be removed by dump trucks. Radio-equipped
flagmen were placed at each end of the single lane area to direct
traffic. The  excavated  shoulders  were  backfilled with gravel
immediately after  a section was confirmed clean. No mishaps
occurred during the excavation of the road shoulders.
Equipment Utilized
  The following list of heavy equipment (with function descrip-
tion) regularly used during the excavation phase is provided below.

• Cat 215   A large capacity trackhoe, equipped with a 48-in.
            straight-edged bucket to obtain a smooth cut. One 215
            cat was equipped with an articulated bucket suitable
            for cutting at different slope angles.
• Case 580 A rubber tired backhoe equipped with a straight-edged
            bucket.
• Cat 910   An articulated highloader that was primarily used to
            transport  supplies  and materials to the work crews,
            as well as to move concrete.
• Gradeall  An all-terrain forklift to transport bags within the con-
            taminated area of the park to the transfer pad.  This
            unit replaced the dump trucks within the hot zone.
• Cranes    Two to three cranes of varying capacities (18 to  32
            ton) were used to remove bags from the hopper and
            to load and unload bags into dump trucks. Cranes also
            were used to suspend concrete for decontamination.
• Hopper   A custom designed and fabricated aluminum funnel
            on a stand to hold the bulk bags and channel dirt into
            the bags from the backhoe bucket.
• Trucks    Dump trucks were used during earlier excavation
            phases to transport bags within the contaminated area
            of  the park and primarily  within the clean areas.
            Dump trucks also were used  to transport bags of soil
            from the Mahaney and Highway  100 excavation
            activities.
• Cat 936   A high capacity, extending highlift equipped with a
            custom designed and manufactured forks made from
            heavy gauge, 6-in. diameter steel pipe. This machine
            was used to remove full bulk bags from the trucks and
            stack the  bags within the storage structures.
• Bag      This stand was custom designed to hold  a full bag
platform    while the  platform forks of the Cat  936 were placed
            under the bag to life the bag to the top row of the
            stack.

Air Monitoring
  An air monitoring plan was included in the original Mitigation
Plan,  although it was modified extensively by the U.S. EPA Air
Branch Division after the initiation of removal activities.  The
modified plan included the use of six hi-volume samplers stationed
on scaffolding around the perimeter of the Quail Run and Chlanda
properties to monitor off-site migration of dioxin-contaminated
particulates3. Air monitoring activities were initiated in May 1985.
During the first 14 days of hi-volume sampler operations, the new
plan required analyses of all six hi-volume filters to obtain  a data-
base. After the 14 data  points were obtained for each site, the
number of samples submitted for analysis was reduced to include
only samples collected from the upwind and downwind  sites. Two
paniculate samplers also were added at two of the perimeter loca-
tions.  During excavation or concrete cleaning operations within
the perimeter, two excavation area samplers located on ground level
near the activity were also operated to monitor potential worker
exposure levels to airborne TCDD.  All hi-volume filters were
weighed to determine the concentration of particulate matter (TSP
in ug/m3) collected in the air during  the 24 hr sampling period.
Meteorological data, specifically wind speed and direction, was con-
tinuously monitored and recorded on a strip chart to determine
which samples were to be analyzed.
   During  excavation activities at the Mahaney residence, four
perimeter monitors were located around the excavation site. The
upwind and downwind samples were analyzed each work day.
During Highway 100 shoulder excavation, one sampler was located
on each side of the highway near the removal  zone, and the
predominantly downwind sample was analyzed.
   Air monitoring was discontinued when removal activities came
to a stop in November 1985 and were resumed upon  startup of
removal activities in April 1986. The generic QAPP was revised
and implemented. Major changes included: (1) extending the sample
collection period from 24 hr to 48 hr,  (2) eliminating sampling of
the TSP and (3) providing a sensitive receptor monitoring  site in
the monitoring network. Location of the samplers is shown in
Figure 2. Approximately 568 samples were analyzed from May
through November 1985 and 266 samples from April through July
1986.
           O
                          Figure 2
     Location of High-Volume Air Samples at the Quail Run Site

  The running  14 data point averages were compared to a 3.0
pg/m3 action limit and a 5.5  pg/m3 No Observed Affect Level
(NOEL). At no time during the project was there a violation of
the 3.0  pg/m3 action  limit by the 14 data point average. The
maximum 14 data point average experienced was approximately
2.4  pg/m3  2,3,7,8-TCDD from  site  HOI.  Values  typically
averaged near the 1 pg/m3 level for all perimeter monitoring sites
and  0.5 pg/m3 for the single sensitive receptor site.

Restoration
  Restoration of the Quail Run park began immediately following
the cleanup of the site in  July 1986. Fill dirt was removed from
the clean property adjacent to the storage buildings to bring the
mobile home park to grade. The plans and specifications were coor-
dinated with the Franklin County Planning Commission with new
requirements on spacing of mobile homes and free space. Thirty-
five pads were poured with patios and walkways. The asphalt roads
were curbed and included asphalt drives to the mobile home pads.
All utilities were underground with  overhead lighting along the
main road. The sewage lagoon was cleaned to place the sewer outlet
distribution line and pad. The sludge from the lagoon was sampled,
and  off-site disposal was  coordinated with  the Missouri Depart-
ment of Natural Resources.
  None of the original Quail Run residents returned to the park
                                                                                                SITE REMEDIATION    309

-------
after restoration. Their trailers were returned to designated sites
requested by the owners and hooked up to utilities. Corrections
were made to any items or materials on the trailer not considered
satisfactory. All trailers were returned by Spring 1987 to the owners
with repair of miscellaneous items on some of the trailers continuing
into July 1987. The restoration of the trailers to  the  owner's
expectations was the most difficult facet of the entire cleanup oper-
ation. The mobile home park, renamed Fox Creek,  had six trailers
on the 35 pads as of July 1987.

Funding
   Because of a number of factors (primarily greater vertical and
horizontal extents of contamination, construction of additional
buildings, the county requirement to build a permanent access road,
lack of  access to some property, more stringent cleanup and
sampling procedures for the mobile homes, reconstruction of the
park to current county specifications and wet weather), costs to
complete the project  greatly exceeded  original  estimates. The
original  removal contractor ceiling approved was  for $1,924,300
and was  raised to  $9,306,322 in April 1986. The final  removal
contract project cost will be approximately  $8,700,000. Project
ceilings and  incurred costs are summarized in Table  1.  A break-
down of removal activity costs is presented  in Table 2.
                            Table 1
                Project Ceilings and Incurred Costs
              For the Quail Run Immediate  Removal.
 ERCS
 Comunity Relations
 latter of Contract
 Other Costs
 Subtotal
 Extramural Cleanup
 Contractor  Costs

 UlWtK COSTS

 HOP
 TXT

 subtotal
 DTORAMIRAL

 Region Salary
 Raglan Travel
 Region Indirect
 HQ Salary
 H2 Travel

 Subtotal
 TOTAL PROJECT
    Approved
    Project
    ceiling

$ 9,306,322
S   702,700
$   177,459
$    14,500
$     7,000
$10,207,981
    424,400
    570,700
                       $   995,100
$   262,500
$    72,000
$ 1,050,000
S     4,000
$     4,000

$ 1,392,500
                       812,595,581
   current
   ABCUIt
   Billed

$ 8,266,325
$   698,349
S    93,423
$        0
$        0
$ 9,058,097
    759,796
    126,332
                             $  886,128
    154,164
     51,142
    540,280
     2,164
     2,104
                                                    $   749,854
                             $10,694,079
                                                                        Table 2.
                                                          Estimated Removal Contractor CoiU for
                                                   Individual Activities of (be Quail Run Immediate Removal
                                               Pre-excavation Planning and Management
                                               ftobile Home Decontamination/Restoration
                                               Access Road
                                               Storage Structure*
                                               Excavation
                                               Pressure Hashing
                                               Kahaney
                                               Air Monitoring
                                               Sit* Restoration
                                               Guard  Service (1984 through 1966)
                                               Handling Charge
                                                                              TOTAL
                                                                    $   49,000
                                                                    $1,424,000
                                                                    $  405,000
                                                                    $  861,000
                                                                    $2,979,000
                                                                    $  213,000
                                                                    $  164,000
                                                                    $  606,000
                                                                    $1,000,000
                                                                    $  219,000
                                                                    $  218,000
                                                                                                 $8,138,000
CONCLUSIONS
  Immediate Removal activities at the Quail Run Mobile Manor
and associated sites to address dioxin contamination in habitable
structures and soil took place over a 2-year period at an approxi-
mate total cost of S11,000,000. Approximately 16,000 yd3 of con-
taminated soil were removed from 5 acres of the park, 1,400 ft.
of highway shoulders and the Chlanda and  Mahaney properties.
Decontamination of 28 mobile homes also was conducted and all
real and personal property was restored or replaced. Statistically
based soil sampling was performed to indicate whether a section
of land was contaminated (both pre- and post-excavation). Exten-
sive air and dwelling  sampling also  was performed.

ACKNOWLEDGEMENTS
  The authors would like to thank members of the Region YD US.
EPA Emergency Planning and Response Branch and the Roy F.
Weston/Tetra Tech Technical  Assistance Team contractors for
their efforts  and input expended during site activities  and the
development of this publication. Riedel Environmental Services,
the primary cleanup contractor, also should be recognized  for their
cooperation and contributions to this removal action.

REFERENCES
1.  Exner. J.H. Keffer.  W.J..  GUben.  R.O. and Kinerson, R.R., "A
   Sampling Strategy for  Remedial  Action at Hazardous Waste Sites:
   Cleanup of Soil Contaminated by Tetrachlorodibenzo-p-dioxin" Hn.
   Wastes and Hn. Mai.. 2.  1985, 503.
2.  Kleopfer, R.D., Gerken, M., Carasea, A. and Morey, D.A., "Analyti-
   cal Support During  Remedial  Action at Sites Contaminated with
   2,3,7,8-Tetrachlorodiberuo-p-dioxin" ACS Symposium Series No. 338,
   Solving Hazardous Wasie Problems:  Learning from Dioxins, p. 259,
   1987.
3.  Fairless, B.J., Bates, D.I.. Hudson. J., Kleopfer. R.D., Holloway, T.T.,
   Morey, D.A. and Babb, T.. "Procedures Used to Measure the Amount
   of 2,3,7.8-Teuachlorodibenzo-p-dioxin in  the Ambient  Air near a
   Superfund Site Cleanup Operation" Environ. Sri. and Tech.. 21,1987,
   555.
 310    SITE REMEDIATION

-------
                                     Groundwater Cleanup  At
                                      Selected  Superfund Sites

                                                    Lisa Haiges
                                               Amoco  Corporation
                                                 Tulsa, Oklahoma
                                               Robert Knox, Ph.D.
                                             University of Oklahoma
                                               Norman, Oklahoma
ABSTRACT
  The focus of this study was on groundwater-related remedial
actions at 36 selected Superfund sites. The source of documenta-
tion is the individual RODs where remedial action alternatives have
been selected for implementation. The objectives of this study were
to examine those  sites whose RODs addressed contaminated
groundwater and to identify any shortcomings in the Superfund
remedial action process,  particularly with regard  to  cleanup
standards. Areas of specific concern included the lack of regula-
tory standards and a set of eight criteria fundamental to the evalu-
ation of the Superfund remedial action decision-making process.
An in'depth examination of each criterion will provide insight to
specific areas where Superfund remedial actions may be ineffec-
tive, inconsistent and/or improved.

INTRODUCTION
  The enactment of CERCLA in 1980 was hailed as the beginning
of a new era in environmental legislation. Unlike its major predeces-
sors, the Clean Water Act and RCRA, CERCLA did not establish
general standards or  behavior for the regulated community '.
Rather: CERCLA provided for "...liability,  compensation,
cleanup and emergency response for hazardous substances released
into the environment and the cleanup of inactive hazardous waste
disposal sites," as defined by Public  Law 96-510 2
  Our interest in performing this study developed while reviewing
literature on the Superfund process  and  several RODs, where
remedial action alternatives had been selected for implementation.
It became evident to us that several shortcomings  exist in  the
Superfund program, especially in the area of contaminated ground-
water cleanup. Many of these problems are attributed to Super-
fund's lack of standards for cleanup actions at hazardous  waste
sites. Another implicating factor is that Superfund does not require
that standards in other federal environmental laws be applied to
these cleanups.  In essence, Superfund provides an overall  direc-
tion, incorporated in the NCP, but does not include any cleanup
standards. This lack of governing standards was the main focus
of this study, specifically addressing a set of criteria dependent upon
established standards, in order to develop  and implement a
successful remedial action.

OBJECTIVE OF STUDY
  The objective of this study was to examine those sites whose ROD
addressed contaminated groundwater and to identify any short-
comings in the Superfund remedial action process, particularly with
regard to cleanup standards. Table 1 is  a list of the 36 sites included
in this study. A brief review of the ROD reveals a great deal of
uncertainty, lack of information and inconsistencies surrounding
the implemented actions. Areas of specific concern include the lack
of regulatory standards, and a set of criteria, fundamental to the
evaluation of the  Superfund remedial action decision-making
process, that are dependent upon set standards.
                         Table 1
   Sites Included in Study, Classified by U.S. EPA Regions
         Region I
         Region II
   McKin Site
   Western Sand & Gravel
   Beacon Heights

   Bridgeport Site
   D'Imperio Property
   Pijak Farm Site
   Price Landfill
   Lone Pine Landfill
   Pollution Abatement Services
   Clean Well Field
   Goose Farm
        Region III
        Region



        Region
IV
        Region VI
Tyson's Dump Site
Heleva Landfill Site
Fischer and Porter Site
Harvey Knott Drum Site
Lackawanna Refuse Site

Whitehouse Waste Oil Pits
Biscayne Aquifer Sites

Reilly Tar Site
Old Mill Site
Eau Claire Municipal Well Field
Main Street Well Field
Verona Well Field
Chem-Dyne
Charlevoix Site
Rummer Sanitary Landfill
New Lyme
Acme Solvents

Bayou Bonfouca
Old Inger Site
  The following eight criteria have been selected in order to address
areas unique to each ROD: type of site, types of contaminants,
                                                                                           SITE REMEDIATION    311

-------
cleanup objectives, cleanup alternatives, cleanup standards, length
of cleanup, cleanup histories and cost projections. Each ROD will
be evaluated with respect to the eight criteria. This evaluation will
be followed by the results of the compilations for all 36 RODs.
An in-depth examination of each criterion will provide insight as
to specific areas where Superfund remedial actions may be ineffec-
tive, inconsistent and/or improved.

THE EIGHT CRITERIA
  The  following criteria  and their respective subdivisions were
selected in order to better identify trends, patterns and/or lack of
information, inconsistencies and uncertainties present in each sub-
ject area. Each criterion will be defined and discussed briefly,
followed by the results of the compilations for all 36 sites.

Type of Facility and Brief  History
  This criterion was included to discover any trends or patterns
present in past Superfund sites, thereby providing insight to what
type of facilities or operations most often results in a Superfund
site. The internal subdivisions included landfills, lagoons, dumps,
storage and/or transfer facilities, well  fields, aquifers, result of
operations,  hazardous waste facilities and others.
  The  results indicated the highest percentage of sites occurred  un-
der the combined classifications of dumps, landfills and lagoons
(39%). Well fields and results of an operation had the next highest
percentages of occurrence (14%); while aquifer (11%), storage
and/or transfer facility  (8%), other  (8%) and  Class I  or II
Hazardous Waste  Facility (6%), followed respectively.
  The  results of the classifications conclude that past dumps, land-
fills and lagoons are the major contributors to present day ground-
water contamination and, ultimately, Superfund sites. The majority
of the past dumps, landfills and lagoons were poorly sited, designed
and  operated, resulting in severe  contamination of  present
resources. The Superfund sites exhibit the undesirable and often
irreversible results of a negligent land  disposal operation.

Types  of Contaminants Present
  The types of contaminants have a greater effect on the outcome
of a site than any other factor. The types of contaminants present
dictate the objectives of a ROD, the types of remedial actions
selected, the safe levels of concentration, the time needed to achieve
those levels and the cost of the remedial action. The study examined
the two main constituent divisions 'organics and inorganics' and
their respective subdivisions; volatile organic compounds (VOCs),
trichloroethylene (TCE),  pesticides and heavy metals.
  The  results show organic contamination present at 94% of the
sites. We also concluded that VOCs occurred at 94% of the sites.
TCE was present in  57% of the sites' groundwater, while pesti-
cides were detected at 17%  of the sites. All inorganic contamina-
tion was present in the form of heavy metals and occurred at 49%
of the  sites.
  A brief description of all the contaminants found at the 36 sites
would  be representative of the wide variability in existing standards
and  proposed criteria for the toxic and often carcinogenic pol-
lutants. The variability and lack of data on the potential carcino-
genic effects of pollutants are directly  related  to  the lack  of
established governing cleanup standards. Currently, the amount
of toxicological information available  is  not sufficient to deter-
mine the risk  associated  with the majority of the toxic priority
pollutants, and  thereby cleanup standards may not have been
adequately developed and implemented.

Type of Objectives
  The objectives of each remedial action are included in the RODs.
These objectives commonly address ".. .the mitigation and control
of contamination in the groundwater." The purpose for including
this criterion in the study was to analyze those RODs that address
aquifer restoration.  Aquifer restoration is an  integral part of
restoring a Superfund site to the point of closure. The subdivisions
include those RODs where aquifer restoration was an objective,

312     SITE  REMEDIATION
where aquifer restoration will be addressed in a future ROD or
where aquifer restoration was not considered as an objective but
rather a result of measures taken. Additional subdivisions include
those sites where the objective was to prevent future groundwater
contamination, or whose objectives did not address any ground-
water contamination.
  The results indicated only  14% of the sites addressed aquifer
restoration in the objectives, while 30% of the RODs will address
it in the future. Seventeen percent of the cleanup actions may result
in some aquifer restoration as the result of other measures taken,
while another 17% of the RODs do not address any aquifer res-
toration. Finally, 70% of the RODs stated objectives towards the
prevention of future groundwater migration. The overall results
indicate a trend towards the prevention of future groundwater
migration; however, the majority of RODs did not address aquifer
restoration in the current objectives.
  The previous trends are the direct result of the lack of cleanup
standards. The lack of cleanup standards has placed great limita-
tions on the overall scope of cleanup objectives. Aquifer restora-
tion is an ideal cleanup goal;  however, until toxicological studies
provide the necessary cleanup data, restoration may be completely
unrealistic. The Superfund program will only be as effective as its
goals. Currently, the RODs limit the goals of the cleanup to a
microscopic portion of the overall program, thereby reducing the
potential effectiveness  of the program.  Until the goals  of the
individual  RODs are  broadened, the results will  continue to
represent only minor accomplishments in terms of Superfund site
cleanups.

Types of Alternatives Chosen
  The types of alternatives chosen for groundwater cleanup meas-
ures are dependent upon the types of contaminants present, the
stated objectives and the most "cost-effective" alternative availa-
ble. The subdivisions included the following: an alternate water
supply,  either temporary or  permanent, and/or a groundwater
extraction/treatment scheme comprised of air stripping, carbon
adsorption, conventional treatment or other. The  RODs imple-
menting groundwater cutoff walls and/or maintenance pumping
were included.
  The results indicate 33% of the sites selected an alternate water
supply for  implementation, with 25% of these sites representing
a temporary measure and 75% providing a new permanent source
of water. Eighty-seven percent of the 36 RODs selected ground-
water extraction and/or treatment for implementation. The most
commonly selected remedies include air stripping and carbon
adsorption (each occurring 35% of the time), while conventional
treatment and other methods were selected for implementation 6%
and 23% of the time,  respectively.  Twenty-three percent of the
RODs had not yet established a specific treatment method. Finally,
58% of the sites utilized maintenance  pumping and 11% of the
sites implemented groundwater cutoff walls to aid in the hydraulic
control  of the contaminant plume.
  The trends reflected in the results indicate that the large majority
of the sites are implementing treatment technologies, although
many of the specific details are deferred to the future. Ideally, there
will be more utilization of new treatment technologies in the future
and advances in the currently implemented methods; however, the
lack of cleanup standards limits the potential advancements in treat-
ment technologies.

Estimated Cleanup  Levels
  This criterion appears to be the focus of many of the uncertainties
and inconsistencies engulfing Superfund remedial actions. The basis
for this  notoriety is the lack of standards governing the majority
of the hazardous waste constituents. The subdivisions were selected
to identify those sites where there was a cleanup level, the cleanup
level was being deferred, there was no cleanup level or a cleanup
level was not applicable.
  The results of this criterion indicated the following: 56% of the

-------
sites had an estimated cleanup level; 11% of the sites deferred a
cleanup level to the future;  14% of the sites had no estimated
cleanup level; and 19% of the sites were not applicable. Based on
the percentages alone,  the results appear very positive with the
majority of sites implementing a cleanup standard; however, the
implemented cleanup standards are incredibly diverse. The RODs
implementing a cleanup  standard identify the inconsistent and
somewhat contradictory selection of cleanup targets. The targets
include groundwater  performance  standards,  drinking  water
standards,  RCRA standards, Maximum  Contaminant  Levels
(MCLs), Alternate Concentration Limits (ACLs), Suggested No
Adverse Response Levels (SNARLs),  background  levels, con-
taminant stabilization and the 10~6 cancer risk level.
  This inconsistency among the cleanup goals is the  direct result
of the lack of cleanup standards. According to the GAO neither
Superfund nor the NCP include explicit standards to be adhered
to when implementing  remedial actions at  hazardous waste sites
3. This absence of cleanup standards has caused great controversy
over what are the actual goals of Superfund remedial actions. The
development of cleanup standards is the ideal solution; however,
the U.S. EPA has been hampered in the past due to the lack of
information on the health effects of exposure to the majority of
toxic constituents  found at  hazardous waste sites (3). Further
progress, according to the GAO, has been limited by funding delays
and staff reductions recommended by the U.S. EPA (4).

Estimated Cleanup Times
  This criterion is directly dependent upon the selected remedial
action  alternatives and  the cleanup target levels and goals. There-
fore, to have an estimated cleanup time, the ROD must state a
cleanup goal. The internal subdivisions are very similar  to the
preceding subdivisions  under the Estimated Cleanup Levels. The
subdivisions include those sites where there is an estimated cleanup
time the cleanup  time  will be  decided in the future, there is no
cleanup time, or a projected cleanup time is not applicable.
  The  results are  as follows: 50% of the sites have an estimated
cleanup time, 14% of the sites deferred a cleanup time to the future,
17% of the sites have no cleanup times and 19% of  the sites are
not applicable. Once again, the majority of sites have an estimat-
ed cleanup time; however, the time necessary to achieve a cleanup
varied  from less than 1  year to greater than 20 years.  The majori-
ty of the sites are  projected to be completed between 1 to 2 years
and 5 years, respectively.  However, the next largest percentage of
sites have a projected cleanup time of  11 years.
  It should be  noted the estimated cleanup times are strictly a
projection based on many assumptions at the time of  the prepara-
tion of the  ROD  and refer to  the time elapsed following the
implementation of a treatment scheme. The implementation of a
treatment scheme may take years to occur,  pending on the design
of the treatment scheme and its dependence on designated cleanup
standards. The cleanup  times are also subjected to extensions
resulting from a provision in the ROD to reevaluate the treatment
system, the cleanup goals and the time necessary to achieve the
desired results.
Date of Discovery and Following Events
  This  criterion was included to examine the amount of time
elapsed from the date of discovery of the site to its current  Super-
fund status. The study was expanded to include the projected date
of site  closure when available; thereby identifying the "cradle to
grave" study of each site. The subdivisions include the current time
elapsed from date of discovery of ROD, the projected site closure
and the site's actual date of discovery (prior to 1975, 1975-1980
and following 1980).
  The  results indicate the time elapsed since the date  of discovery
of a site to the current ROD varies from 1 to 17 years, with the
greatest percentage of the RODs occurring at 4 and 5 years,  respec-
tively,  following the date of discovery. This elapsed time period
may consist of cease and desist orders, legal actions, immediate
removal actions,  feasibility studies (FS), remedial investigations
(RI), approval of RI/FS, initial remedial measures and, ultimately,
a Record of Decision. The average time elapsed since the date of
discovery of a site to the project cleanup or final site closure was
29 years, indicating that the average life span of a Superfund
hazardous waste site approaches 30 years. Finally, a brief exam-
ination of  the actual date of discovery  of a site indicated  the
following:  those sites discovered prior to 1975 had  an average
response time of 11.7 years; those sites discovered during 1975 to
1980 had an average of 7.2 years elapse prior to a ROD; and those
sites discovered  proceeding the enactment of CERCLA had an
average response time of 3.7 years.
  The sites discovered following 1980 have a  reduced average
response time. However, the cost of a hazardous waste site is greatly
increased with every action and time delay; thereby indicating the
possibly unnecessary and often outrageous costs of the cleanup
program. The effectiveness of the Superfund program would be
greatly increased with an overall reduction  in the time, elapsed since
the date of discovery of these Superfund sites.

Costs of Remedial  Actions
  This criterion, examining the costs of remedial actions !4 was
included for the following two reasons: (1) to give the reader an
idea of the incredible costs  of each remedial action and (2) to
determine whether the chosen alternative  was the minimum cost,
intermediate cost or  the  maximum  cost alternative. The cost
criterion is  one of the main sources of criticism encompassing  the
Superfund  program. The $9 billion "Superfund" is to be  spent
on the remedial action cleanups and ultimately result in final site
closures of the cleaned up sites.
  The results reveal the total capital costs of the cleanup program
range from $262,000 to $57,672,000. The average total capital cost
of remedial actions is $5,619,394, with the largest percentage of
sites occurring in the $1 million cost range. The next largest per-
centage of  sites is in the $3 million cost range.  The addition of
the operation and maintenance (O&M) costs may increase the overall
cost of the remedial action by over 50%. The potential is  great
for the overall costs to nearly  double the original capital  cost,
thereby supporting the theory that  the  best way to  attack a
politically popular cause is to throw money at it.
  However, it is important to determine whether the most "cost-
effective" remedial action alternative is "cost-dependent." A cost-
effective remedial action is defined in the NCP as "the lowest-cost
alternative  that  is  technically feasible and reliable and which
effectively mitigates and minimizes damages to and provides ade-
quate protection of public  health, welfare, or the environment."
The results  indicated 33% of the RODs selected the least-cost
alternative; 46% of the RODs selected the  intermediate-cost  alter-
native; and 21 %  of the RODs selected the highest-cost alternative.
The majority of the  RODs have selected  an intermediate-cost
alternative  for implementation, indicating the selection of  alter-
natives in this study was not entirely cost dependent.
  A final discussion on the cost criterion addresses the criterion's
dependence on established cleanup standards. Without standards,
the cleanups result in many ineffective, repetitive, unsure actions,
while the results may not even be socially or politically accepta-
ble. The total costs  of the cleanups could  be reduced through the
limitation of the number of phases, operable units and remedial
actions implemented at the sites. The development of toxicological
studies and resulting data could greatly decrease the necessary time
and the expenditures of funds, while immensely increasing the over-
all success  of the Superfund program.

CONCLUSION
  The results of this study indicate that a  great number of short-
comings exist  in the Superfund process.  Several of these short-
comings were identified while examining the eight criteria.  Each
criterion is  fundamental to the Superfund process; however,  the
results indicate a great lack of information, inconsistencies, and
uncertainties present in each study area.
                                                                                                 SITE REMEDIATION    313

-------
  Based on the results of the study, the following conclusions can
be drawn:

• Great care must be taken to ensure that current land disposal
  methods will not result in future Superfund sites. Through the
  use of impermeable liners and caps, barriers, leachate collection
  systems, surface water diversions and other methods, the risk
  of groundwater contamination may be greatly reduced. However,
  all man-made attempts to control the surface and subsurface are
  susceptible to failure over time.
• The Results of the Types of Contaminants  indicate the  great
  majority of the sites are contaminated with VOCs. Fifty-seven
  percent of the sites  are contaminated with TCE, a suspected
  human carcinogen. The majority of the contaminants detected
  at the  36  sites are suspected or known carcinogens; however,
  the U.S.  EPA's policy is that there is no scientific basis for
  estimating safe levels for carcinogens'. Currently, the amount
  of toxicological information available is not sufficient to deter-
  mine the risk associated with the majority of the toxic priority
  pollutants.  This lack of data on the carcinogenic effects of
  pollutants is directly related to the lack of set cleanup standards;
  as  a result, cleanup standards  may  not be  developed and
  implemented.
• The majority of objectives addressed the prevention of future
  contamination; however, only 14% of the sites addressed aquifer
  restoration. The prevention of future releases of contaminants
  is extremely important, but what about the existing contamina-
  tion? What is the  benefit of preventing future contamination,
  when the  groundwater is currently non'usable? More emphasis
  needs to be placed on the cleanup and restoration of the con-
  taminated aquifers. By concentrating solely on future contami-
  nation, the life of a site under Superfund status will  perpetuate
  indefinitely.
• The majority of the types of alternatives selected for  implemen-
  tation  included a groundwater extraction  and/or treatment
  scheme; however,  very few of the sites utilized  state-of-the-art
  technologies. Ideally, Superfund should be promoting high-tech
  treatment methods, yet the results indicate continued use of air
  stripping or carbon adsorption. One of the 36 sites incorporated
  a biological disc in its selected alternatives,  and only one site
  proposed  and selected a carbon adsorption-air stripping treat-
  ment train. The advancements in treatment technologies are
  necessary to improve the acceptable treatment levels, thereby
  requiring  the technologies to push  the state-of-the-art.
• The implementation of  extremely  diverse  and inconsistent
  cleanup standards has resulted in ineffective site cleanups. The
  selected standards ranged from background  levels to contaminant
  level stabilization. The inconsistency among the cleanup  goals
  is the direct result of no set cleanup standards and continues to
  be a focal  point of shortcomings in the Superfund program. Until
  such standards are developed and implemented, the  Superfund
  program will continue to have a very low success rate.
• The estimated cleanup times indicated an average of 1 to 5 years
  to treat the contaminated groundwater. However, the cleanup
  times are  subject to extensions,  and without set cleanup stan-
  dards, are only hypothetical  projections that may  ultimately
  result in the passage of many years prior to achieving site cleanup.
  Until set standards are discovered and implemented, the esti-
  mated cleanup times will continue to fall prey to deferrals and
  reevaluations resulting in undesirable and possibly lengthy, ex-
  tensions.
• The time elapsed since the date of discovery of a site is the direct
  result of the number of deferrals. The continuous deferral of
  actions is the result  of "tunnel vision" within the Superfund
  process. The U.S. EPA approaches each site by addressing one
  problem or aspect at a time, thereby resulting in the expenditure
  of millions of dollars on small-scale  investigations, studies and
  responses. The  "tunnel   vision"  or  deferrals range  from
  addressing aquifer restoration to designing selected treatment
  schemes,  and selecting target levels, to reevaluating any pre-
  viously implemented decision.
• Finally, the costs of the remedial actions indicate a gross expen-
  diture of funds on little or no measured results. The current NPL
  indicates thai only four of the 700 sites have ceased implemen-
  tation activities, with  no  mention if the  four sites have been
  restored 6. The costs of the remedial actions are directly depen-
  dent on the  lack  of cleanup  standards, "tunnel vision" and
  deferrals, resulting in many ineffective, repetitive and often un-
  acceptable actions and results.

  The future success of the Superfund program is dependent upon
the development of set cleanup standards. However, standards de-
velopment will require extensive toxicological studies and data
which, in the past, were subject to funding delays and reductions
'. An  alternate solution to the current Superfund program would
place  greater emphasis on these toxicological  studies through
increased funding and  qualified personnel.
  This study identified several specific  areas where shortcomings
originate in the Superfund program; however, the majority of the
cases were the result of a  lack of cleanup standards. Through the
development and implementation of cleanup standards, the over-
all Superfund program would be greatly  improved resulting in more
consistent and effective total site cleanups.

REFERENCES

1. Lowrance, S. and Stanley. E., "Directions in Hazardous Waste Cleanup:
  Choice in Superfund Implemenn'on," Proc. of the Fourth National Con-
  ference on  Management  of  Uncontrolled Hazardous Waste Sites,
  Washington, D.C..  1983, 1-4.
2. Public Law 96-510. CERCLA, 1980.
3. U.S. General Accounting Office, "Cleaning Up Hazardous Wastes: An
  Overview of Superfund Reauthorization  Issues," GAO/RCED-85-69,
  U.S. General Accounting Office, Washington, D.C, March 1985.
4. U.S. General Accounting Office, "Clearer EPA Superfund Program
  Policies Should Improve Cleanup Efforts," GAO/RCED-83-54, Report
  to Controller General. Washington,  D.C., February 6, 1985.
5. Sinig, M., Priority  Toxic Pollutants, Noyes Data Corporation, Park
  Ridge, NJ,  1980.
6. 40 CFR 199-399, 1986.
 314    SITE REMEDIATION

-------
                               Pilot-Scale  Bioremediation  at  the
                                   Brio  Refining  Superfund  Site

                                                     Bruce  S. Yare
                                           Monsanto Chemical Company
                                                  St. Louis,  Missouri
                                                  Derek Ross, Ph.D.
                                                  Ecova Corporation
                                               Redmond,  Washington
                                                  David W.  Ashcom
                                                  Ecova Corporation
                                               Redmond,  Washington
ABSTRACT
  A pilot-scale, solid-phase air stripping and biological treatment
facility was constructed at the Brio Refining Superfund Site in order
to demonstrate the feasibility of bioremediating contaminated soils
and organic residues. The site, located near Friendswood, Texas,
has a large volume of soils containing styrene, still bottom tars
and chlorinated hydrocarbon solvents.
  The treatment facility consisted of an enclosed, lined treatment
bed containing 200 yd3 of contaminated soil from one of the
backfilled storage lagoons located at the site. The liner was an 80
mil HDPE synthetic membrane with heat-welded seams. A sand
drainage layer was placed on top of the liner and a 6-in thick layer
of contaminated soil was placed on top of the sand. Nutrients and
inoculum were applied to the treatment bed through an overhead
spray system. The treatment bed was tilled daily to increase soil
surface area and provide aeration. Volatile emissions from the treat-
ment bed were contained by a plastic-film greenhouse and routed
to carbon adsorption units.
  Sampling after 21 days of operation indicated that greater than
99% of the volatiles present  in  the contaminated soil had been
removed by air stripping. Samples collected after 94 days of
operation demonstrated that 89% of the semi-volatile compounds
were degraded. Pheneathrene, the primary semi-volatile organic
compound found on site, had a half-life of 33 days in the study.
This was a significant improvement in degradation rate over the
69 to 298 day half-lives reported in the literature.
  This pilot-scale treatment  facility, constructed at the site and
operated under field conditions, effectively demonstrated an effi-
cient, cost-effective process for remediating organic compounds
found in on-site soils.

INTRODUCTION
  The Brio Refining Superfund Site is located in Harris County
approximately 20 miles southeast of Houston, Texas. The 588acre
site is about 1.5 miles southwest of the Ellington Field exit off
Interstate Highway 45 South (Gulf Freeway). Friendswood, Texas
is located  approximately 1 to 2 miles to the  south. The site is
bounded on the north by the Southbend subdivision, on the south
and east by pasture land and oil fields and on the west by Dixie
Oil Processors, an associated Superfund site. Prior to 1956, the
site was part of the Friendswood oil field. Diked oil stock tanks,
lagoons and other structures associated with crude oil production
and storage were located on  site.
  From  1957 to 1969, the major process operations at the Brio
site were the regeneration of copper catalysts and reprocessing of
styrene and vinyl chloride production residues. Copper catalyst
regeneration was discontinued in 1970. Between 1969 and 1972,
another process operation used hydrogen sulfide blended with spent
caustic to produce cresylic acid,  sodium sulfide and  sodium
cresyllite. Operations during 1972 to  1975 are not well documented.
However, styrene tar reprocessing apparently continued through-
out this period. From 1975 to 1978, styrene tar, off-spec diesel fuel,
ethylbenzene, phenol bottoms, cutter stock, caustic, crude oil, blend
oil, polyethylbenzene bottoms and crankcase oil feedstocks were
used to  produce aromatic oil, fuel oil,  ethylbenzene,  toluene,
cumene, sodium sulfide, creosote extender and 50% caustic.
  The recycling and recovery plant at the Brio Refining Site was
converted to jet fuel production in 1978. Jet fuel, diesel fuel, resi-
dual oil,  naphtha, kerosene and fuel gas were produced by dis-
tilling crude oil. Jet fuel production continued until 1982 when all
operations at the site were discontinued. The site has remained
inactive  since then.
  The site was nominated  for inclusion on the NPL primarily on
the basis of incomplete closure of raw material storage pits. There
are 21 raw material storage pits at the Brio Refining site.  The first
pit closures were completed during 1969 and 1970. Seven additional
pits were closed between 1972 and  1975. Another four pits were
closed in 1976 and 1977. The final on-site pit was closed in 1979
or 1980. Closures were conducted by mixing soil, sand or calcined
clay with the residues left in place. Soil cover was placed over the
stabilized pit  material.
  The RI/FS, conducted by the PRPs, was started early in  1984
and will be finished late in 1987. Solid-phase biodegradation was
one of the potential remedial technologies evaluated during the
Feasibility Study. Rather than relying on literature reports or bench-
scale studies, the Brio Task Force decided to undertake an on-site,
pilot-scale biodegradation  demonstration to establish conclusively
that bioremediation was capable of destroying  the organic com-
pounds  found at the Brio Refining site.
  The biodegradability of the pit backfill material was determined
by Microtox testing. Of the 11 pits tested, two were found suit-
able for biodegradation without dilution. Of these two pits, the
one with the lowest concentration of volatile organic compounds
was selected as the source of material for the pilot-scale biodegra-
dation demonstration. This pit was designated as Pit 0 during the
RI/FS. Additional samples of the Pit 0 backfill were collected in
order to conduct a bench-scale evaluation of the biodegradability
of the organic compounds present in the backfill. This testing
indicated that the ketones, short-chain chlorinated hydrocarbons,
chlorinated aromatic hydrocarbons and aromatic hydrocarbons
found in the pit backfill could be removed by air stripping or bio-
logically destroyed by  indigenous microorganisms. On this basis,
                                                                                              SITE REMEDIATION     315

-------
the Brio Task Force decided to undertake a pilot-scale demonstra-
tion of biodegradation of backfill material from Pit 0.

TREATMENT FACILITY CONSTRUCTION
AND OPERATION
  Treatment Facility Construction
  The pilot-scale, solid-phase biological treatment facility consisted
of a plastic film greenhouse enclosure; a lined soil treatment bed
with an underdrain; an overhead spray system for distributing
water, nutrients and inocula; an organic vapor control system; and
a fermentation vessel for preparing inoculum and/treating leachate.
  Two standard plastic film greenhouses were erected side by side
at the site of the pilot-scale treatment facility to contain any vapor
emissions from the treatment bed and eliminate the need to control
or treat rainwater (Fig 1).  Once the greenhouses were erected, the
soil treatment bed was constructed. The soil treatment bed con-
sisted  of a permeable  sand underdrain  overlying a synthetic
membrane liner. The underdrain  was a layer of bank sand ranging
in thickness from 6 in at the center of the treatment bed to 18 in
at its outer edge. The upper surface of the underdrain formed a
nearly level surface on which the Pit 0 soil was applied. The leachate
collection system, designed for a 20 gal/min inflow, included two
lateral perforated pipe drains and a 140-gaI capacity gravity-flow
sump. Leachate was pumped from the collection sump to the on-
site fermenter vessel for treatment. The synthetic membrane liner
was constructed of 80-mil high-density polyethylene with  heat-
welded seams. The edges of  the liner were secured to the inside
of the greenhouse structural  arches with self tapping screws.
                     - PLASTIC HUH GREENHOUSES
                                              MOW UNEM •
                           Figure 1
         Plot-Scale Solid Phase Biological Treatment System
  The overhead spray system for distributing water, nutrients and
inocula consisted of four individual headers, one for each of the
four treatment lanes. The spray pattern was designed to eliminate
overspray from one treatment lane to another. Water, nutrient or
inocula mixtures were applied through 32 individual spray heads
per treatment lane. Each spray head was capable of supplying 0.79
gal/min at  30 lb/in2.
  Organic  vapor emissions were controlled  by adsorption on
activated carbon. Air was removed  from  the treatment facility by
three 3,000 ftVmin fans and sent through three activated carbon
absorbers. Each of the carbon absorbers was capable of treating
an air flow of 3,000 ftVmin.
  Following construction of the treatment facility, approximately
200 yd3 of  soil were removed from Pit 0 and transferred to the
treatment facility. Soil  was excavated with a 12 ft reach backhoe
and transported to the treatment bed using two tracked front  end
loaders. The excavated soil was placed on top of the prepared treat-
ment bed. Due to the cohesiveness of the clay soil, the pit backfill
material was allowed to dry before final grading. For several days
the tracked front end loaders were run back and forth over the
pit backfill to break up large blocks of soil and distribute material

316     SITE REMEDIATION
                                                                  evenly over the treatment bed. After 3 days of manipulation, the
                                                                  clay was amenable to tillage by a power rototiller attached to a
                                                                  tractor. Soil moisture content was low enough after 6 days to al-
                                                                  low addition of nutrients.

                                                                  Treatment Lanes
                                                                    The soil treatment bed was divided into four lanes so that dif-
                                                                  ferent methods of optimizing microbial activity and biodegradation
                                                                  rates could be evaluated (Fig 2). A control lane, which received
                                                                  only tilling and water additions, was established to provide a base-
                                                                  line to evaluate the effectiveness of the following three treatment
                                                                  processes:  (1) nutrient addition, (2) single microbial inoculation,
                                                                  and (3) multiple microbial inoculations.
                                                                                                   TREATMENT LANES
                                                                                              Figure 2
                                                                                  Biological Treatment Facility Layout
                                                                      The nutrient-adjusted lane was sprayed with inorganic nutrients
                                                                    (nitrogen and phosphorous) to stimulate the activity of indigenous
                                                                    microorganisms in the contaminated soil. This treatment process
                                                                    was designed to determine the rate at which the indigenous micro-
                                                                    bial population could degrade  organic compounds with only the
                                                                    addition  of oxygen and inorganic nutrients. The single inocula-
                                                                    tion lane was sprayed once at  the start of operations with inor-
                                                                    ganic nutrients and a high concentration of microorganisms isolated
                                                                    from Pit 0. This treatment process  was designed to determine if
                                                                    the rate of organic compound degradation could be increased by
                                                                    augmenting the existing microbial activity from Pit 0. After an
                                                                    initial inoculation, the multiple inoculation lane was subsequently
                                                                    sprayed with nutrients and inoculum at approximately  10  day
                                                                    intervals. The inoculum was developed from soil removed from
                                                                    the lane and water from the leachate collection system. This treat-
                                                                    ment was designed to determine if the rate of removal of organic
                                                                    constituents could be accelerated by increasing the frequency of
                                                                    application of microorganisms.

-------
Treatment Facility Operation
  The pilot-scale treatment facility was operated for 94 days. The
soil in the treatment facility was tilled daily to optimize contact
between microorganisms and the organic constituents present in
the pit backfill material and  to  ensure adequate aeration  for
microbial activity. Tilling also facilitated the air stripping of vola-
tile organic compounds. Soil moisture content,  soil temperature
and soil pH were monitored to ensure that they remained within
ranges conducive to microbial activity. Water, nutrients and inocula
were added as required to the treatment bed through the overhead
spray system.

Sample Collection and Analysis
  Soil samples were collected on January 26 (Day 0), February 16
(Day 21), March 25 (Day 58) and April 30 (Day 94),  1987. Each
of the four treatment lanes was divided into four quadrants (Fig 2).
These quadrants were in turn subdivided into three 10 ft by 10 ft
sampling cells. During each sampling episode, one sample was col-
lected from each of the 16 quadrants. Samples from each quadrant
were a composite of three subsamples,  one from each of the three
sampling cells in a quadrant. The subsample locations were selected
randomly.
  Soil samples were analyzed for volatile  and semi-volatile organic
compounds in order to determine the  rate of organic compound
degradation and measure the effectiveness of the three treatment
processes. In addition, the soil samples were analyzed for soluble
ammonium, nitrate and phosphate to determine if the concentra-
tions of these nutrients were sufficient to ensure maximum micro-
bial growth and organic compound degradation.
  Microbial growth was monitored using standard plate count tech-
niques.  Radiotracer analyses were  done to confirm biological
mineralization of organic compounds. Soil samples were spiked
with 14C-labeled glucose and phenanthrene and the  amount of
these compounds biologically  converted to  14C-labeled carbon
dioxide was measured. The 14C-labeled glucose mineralization was
used to monitor general microbial activity. Radiolabeled phenan-
threne was used to prove biological mineralization of site specific
compounds and demonstrate stimulation of phenanthrene-specific
degraders.
 RESULTS
 Microbial Activity
   The initial concentrations of microorganisms in the pilot test area
 ranged from 104 to 107 colony forming units per gram wet weight
 of soil (cfu/g). During the operation of the pilot-scale treatment
 facility, these numbers increased to 107 to 108 cfu/gm (Table 1).
 The increase and eventual stabilization of the numbers of micro-
 organisms in the treatment bed indicated that stable microbial com-
 munities were present.
                           Table 1
       Number of Aerobic Heterotrophic Organisms Present
          In Treatment Lanes. Pilot-Scale Bioremediation,
             Brio Refining Site, Friendswood, Texas


                     	Nuaber of Microorganisms (Cells/Kg)
     Lane
  Control
 Nutrient
 Adjusted

 Single
 Inoculation

 Multiple
 Inoculation
                     Day 0
                    1.78x10"
                    7.49xlOJ
2.92x10'
2.76x10"
                               1.70x10'
                               4.83x10'
           7.13x10'
           6.89x10'
                                          3.03x10
                      5.46*10'
3.33x10'
4.93x10'
                                                     9.69xlOJ
                                                     8.40x10
                                  4.66x10ฐ
                                  5.27x10
                                                Removal of Volatile Organic Compounds
                                                  The predominant volatile organic compounds detected in the Pit
                                                0 backfill material placed in the treatment facility were ethylben-
                                                zene, styrene and toluene. These compounds were detected at max-
                                                imum  concentrations  of 4,400,000,  240,000,  and  510,000,
                                                respectively. Methylene chloride and 1,1,2-trichloroethane were also
                                                detected but at lower concentrations. Methylene chloride concen-
                                                trations ranged from 530 to 20,000, while 1,1,2-trichloroethane con-
                                                centrations ranged  from 520 ppb  to 110,000 ppb. Acetone;
                                                2-butanone; chlorobenzene; 1,1-dichloroethane; methylene chlo-
                                                ride; 1,1,2,2-tetrachloroethane, 1,1,2-trichloroethane; andxylene
                                                were detected at concentrations ranging from 3,100 to 88,000; 3,700
                                                to 54,000 ppb; 3,400 to 26,000 ppb; 2,300 to 200,000 ppb; 530 to
                                                20,000 ppb; 4,000 to 5,100 ppb; 520 to 110,000 ppb  and 550 to
                                                180,000 ppb,  respectively.
                                                  The  concentrations of the volatile organic  compounds in  the
                                                treatment facility were reduced by more than 99% over the 94-day
                                                period of operation (Table 2). Most of this reduction occurred
                                                within the first 21 days of operation and was predominantly due
                                                to air stripping. Volatile compounds of both high and low volatility
                                                were removed with equal efficiency. For example, the concentra-
                                                tions of methylene chloride and 1,1,2-trichloroethane, both highly
                                                volatile compounds, were reduced by more than 99%. The con-
                                                centrations of ethylbenzene  and styrene,  both  low volatility
                                                compounds, also were reduced by more than 99%.
                                                                          Table 2
                                                             Volatile Organic Compound Removal.
                                                                  Pilot-Scale Bioremediation,
                                                                      Brio Refining Site,
                                                                     Friendswood, Texas

                                                                	Total Volatile Organic* (PPB)
                                                 Control
                                                                     Nutrient
                                                                     Adjusted
                                                                Day 0    Day 21    Pay 58


                                                               25.972      81        17
                                                 Single
                                                 Inoculation

                                                 Multiple
                                                 Inoculation
                                         39,460      40        14        12       99.90*


                                        273,184      13        16        25       99.99*


                                        101,868      10        19        27       99.9*


                          *Reductlon After 21 Days of Operation
  Two methods were used to estimate the amount of volatile
organic compounds removed from the contaminated soils by air
stripping: (1) concentration of volatile compounds absorbed in the
activated carbon units and (2) air emission data collected during
facility operation. The amount of volatile compounds air stripped
from  the contaminated soils ranged from  137  kg to 159  kg,  a
removal rate of approximately 7 kg/day.

Degradation of Semi-Volatile Organic Compounds
  Phenanthrene was the predominant semi-volatile organic com-
pound detected in the Pit 0 backfill material placed in the treat-
ment  facility. Phenanthrene concentrations ranged from 440 to
170,000 ppb and the average phenanthrene  concentration was
36,300 ppb. 2-Methylnaphthalene  and naphthalene were also
detected in the contaminated soil. 2-Methylnaphthalene concen-
trations ranged from 6,200 to 170,000 ppb, with an average con-
centration of 50,700 ppb. Naphthalene concentrations ranged from
130 to 96,000 ppb and the average concentration was 19,500 ppb.
Over the 94-day operation of the pilot-scale biological treatment
facility,  semi-volatile organic compound  concentrations were
reduced an average of 89% (Table 3).
                                                                                                  SITE REMEDIATION    317

-------
                           Table 3
          Semi-Volatile Organic Compound Degradation.
                   Pilot-Scale Bloremedlation,
                      Brio Refining Site,
                      Friendswood, Texas

               Totil JCTl-Vol.tll. On.nl c Compound.  (PPป)
Un.
Control
Nutrlant
Adjuat.d
Slngl.
Inoculation
Multipla
Inoculation
Day Q
18,900
16,100
56,98)
16,496
Day 81
9,3*6
6,999
4,610
6,028
D.T 58
6.078
5.325
3,967
6,611
Day 9*
2.928
1.402
2.023
2,800
Kaduct ton
84.511
91.291
96.491
83.031
Phenanthrene Degradation
  Due to its predominance in the contaminated soil from Pit 0,
phenanthrene was used to determine the effect of the various treat-
ment processes on the degradation rate of semi-volatile organic
compounds. Over the 94 days of facility operation, phenanthrene
concentrations were reduced an average of 84% (Table 4). During
the first 21 days of operation, phenanthrene degradation occurred
at a relatively rapid rate (Fig 3). For the remainder of the demon-
stration project the phenanthrene degradation rate was approxi-
mately 124 mg/ki/day. At this degradation rate, approximately
131 days would be required for the phenanthrene concentration
to reach 330 ppb, the analytical detection limit using the U.S. EPA
approved procedure.
                           Table 4
                  Phenanthrene Degradation.
                  Pilot-Scale Bloremedlation,
                      Brio Refining Site,
                     Friendswood, Texaa
                 Ph.nanthr.na Concantratlon (PPi)
  Phenanthrene half-life values for the control, nutrient-adjusted,
single inoculation and multiple-inoculated lanes were 40.8, 33.0,
25.7 and 43.3 days, respectively (Table 4). A statistical analysis
of the data demonstrated that there was no significant difference
in the rate of phenanthrene degradation in the different treatment
lanes. The initial phenanthrene concentration apparently was the
parameter controlling  the  rate  of phenanthrene  degradation
(Table 5). The data collected during this demonstration project
suggest that  aeration and the amount of contact between the
microorganisms and the contaminated soil also were parameters
that governed the rate of phenanthrene degradation.
                                                                                              Table 5
                                                                       Effect of Initial Concentration on Phenantnrene Degradation.
                                                                                      Pilot-Scale Bloremedlation,
                                                                                          Brio Refining Site,
                                                                                         Friendswood, Texas
 Initial Concentration,  PPB
      1,000
      5,000
4,999
9,999
     10,000    -  49,999
     50,000    - 100,000
    greater  than 100,000
Average  Reduction, I


          27.4


          33.4


          67.2


          94.0


          96.7
                                                                                               Table 6
                                                                        Radio Tracer Demonstration of Phenanthrent Blodegradattoa.
Initial Final Bait Ufa
Lan* (D.y 0) (Day~54) Udoctlon (Daja)
Control 27.8SO 5.725 79.441 40.8
•utrlant 19.400 2,712 86.02Z 33 0
Adju.t.d
Inoculation
Hultlpl. 24,360 5.27S 78.351 43.3 Control
Inoculation
Nutrl.nt
100 -i MJuatao1
- Slnซl.
J Inoculation
Z 80 " T Multlpla
S Inoculation
<
ฃ 60 -
111
u
u
u 40 •
u UM
| 20- >s. T Control
ฃ ' •* 	 . Nutrlant
" Adjuit.d
0 •
Slngla
1 ' i i Inoculation
0 21 88 04
TIME (DAYS) Multiple
_. _ Inoculation
Figure 3
Reduction in Phpnanthrpn* Pnnr^ntratinn All Counta Corracti
nioi-scaje Hiortmeatanon,
Brio Refining Site,
Friendswood, Texas
14C-Clucoaซ Hlaaralliatlon,
D.T 0 Oar 11 D.T 58
33.0 4J.3 25.3
17.7 47.3 30.4
10.6 44.* 29.1
24.1 45.6 35.2
C-rh.ninthr.nl Mln.rallr.atlo
D.T 0 D.T 21 Par 38
19.0 42.0 42.6
22.5 51.2 42.4
11-6 36.0 43.3
10.7 46.1 43.3

P.rcaat
Day 94
9.9
21.0
17.8
15.9
or f.rc.nt
Daj 94
42.6
46.7
43.8
42.8
318    SITE REMEDIATION

-------
  Since there was no significant difference in the rate of phenan-
threne degradation observed in the different treatment lanes, all
of the data were pooled to determine the rate of phenanthrene bio-
degradation in the treatment facility. The average half-life value
was 33  days, significantly less than reported half-life values of 69
to 298  days in other solid-phase biodegradation systems.
  The initial percent mineralization of 14C-labelled glucose to 14C-
labelled carbon dioxide within the pilot test area was low and
extremely variable.  During the first 21 days of operation of the
treatment facility, the percent mineralization stabilized at approx-
imately 46% and subsequently decreased to approximately 11%
by day 94 (Table 6). The initial 14C-labelled carbon dioxide per-
cent mineralization of I4C-labelled phenanthrene to ranged  from
0 to 51.5%. During the 94-day operation period, the percent miner-
alization increased to,  and stabilized at, approximately 44%.
  The initial increase and stabilization in the percent mineralization
of 14C-glucose and 14C-phenanthrene indicates that the pilot test
operations stimulated and  promoted an  even distribution of
microbial activity, phenanthrene biodegradation potential in par-
ticular, within the treatment bed. The subsequent decrease in 14C-
glucose  activity suggests that facility operations promoted the
establishment of  a  microbial  population that  preferentially
degraded phenanthrene,  the most  predominant semi-volatile
organic constituent in the soil.


CONCLUSIONS
  The pilot-scale biological  treatment facility constructed at the
Brio Refining Superfund Site conclusively demonstrated that target
compounds such as  1,2-dichloroethane, 1,1,2-trichloroethane and
phenanthrene could be removed effectively from soils using an on-
site treatment technology other than incineration. The process
removed volatile organic compounds by air stripping and destroyed
semi-volatile organic compounds by biodegradation.
                                                                                                   SITE REMEDIATION    319

-------
         Evaluation  of  Well Field  Contamination  Using  Downhole
                    Geophysical  Logs  and  Depth-Specific  Samples
                                                   George T. Ring
                                                   Thomas C. Sale
                                                    CH2M HILL
                                                  Denver, Colorado
ABSTRACT
  Downhole geophysical logs and depth-specific samples collected
during an aquifer test may be used to characterize the vertical
nature, extent and migration of contaminants in an aquifer. This
approach was taken at the San Jose 6 Superfund site, Albuquerque,
New Mexico, as an alternative to a program of installing deep
monitoring wells.
  Vertical velocity and brine injection logs were used to define
vertical flow between hydrologic units under static conditions
through the San Jose 6 well bore. In addition, vertical velocity logs
were used to define the percent contribution  to the well of each
screened contributing hydrologic unit under pumping conditions
of 1,000,1,500 and 1,800 gal/min. Temperature and resistivity logs
were used to further define flows in the well bore under static and
pumping conditions.
  Depth-specific sampling was conducted under pumping condi-
tions of 1,800 gal/min at 190, 230, 330, 490, 550. 660 and 730 ft.
By combining the water quality data with the  flow data from the
velocity  logs, the contamination concentration present in each
contributing hydrologic unit was calculated using a mass balance
calculation.
  Contaminant migration pathways to San Jose 6 were evaluated
by recording water levels  in 30 monitoring wells across the site.
Changes in head across the site were used to evaluate vertical
leakage from the contaminated unit above the screened portion
of Well San Jose 6.
  Final results from the aquifer test delineate the vertical nature
and extent of contaminants in  the vicinity of Well San Jose 6 and
define the  mechanism by which contaminants have reached the
aquifers from which Well San Jose 6 produces water.

INTRODUCTION
  An aquifer test was conducted at Albuquerque's Municipal Well
San Jose 6 as part of the remedial investigation for the San Jose
6 Superfund site, Albuquerque, New Mexico. During the aquifer
test, downhole logs  and depth-specific samples were collected to
define the  vertical nature, extent and migration of groundwater
contaminants at the site.  This paper describes the collection of
temperature, resistivity, spinner and brine injection logs, and depth-
specific samples during the  San Jose aquifer  test to characterize
site contamination.
  Well San Jose 6 is part of the Albuquerque  municipal San Jose
well field. In 1980 a variety of volatile organic  compounds (VOCs)
were detected in groundwater produced from  San Jose 6, This led
to removing the well from service and including the San Jose 6
site on the NPL.
  Groundwater contamination at the San Jose well field is of par-
ticular concern because of Albuquerque's sole reliance on ground-
water for water supply. The broad objectives of the aquifer test
and the remedial investigation were to provide sufficient informa-
tion  to evaluate feasible and cost-effective alternatives for rite
remediation.
  The San  Jose 6 site lies within the Rio Grande Rift and it
underlain by the Santa Fe Formation, which consists of uncon-
solidated sands  and gravels  with occasional beds of silts and
clays.1 Water pumped from municipal well San Jose 6 is produced
from the Santa Fe  Formation. Recharge occurs predominantly
through vertical leakage from overlying aquifers and from lateral
inflow of groundwater. Table 1  lists well design/capacity infor-
mation for  San Jose 6.

                          Table  1
           San Jose 6 Well Design/Capacity Information
 Completion Depth
 Screened Interval
 Blank Interval
 Gravel Pack Interval
 Well Capacity
 Motor
 Pump Setting
912 feet
180 to 912 feet
0 to 180 feet
96 to 912 feet
2,300 gallons/minute
ISO horsepower
140 feet
  The aquifer test conducted at the San Jose 6 well was designed
as an alternative to a program of installing monitoring wells to
depths exceeding 800 ft. By collecting downhole geophysical logs
and depth-specific samples, details were obtained regarding the
nature, extent and migration of contaminants in the well vicinity.
Collecting these data using monitoring wells would have been more
costly and time-consuming.

METHODOLOGY
  The downhole work conducted at Well San Jose 6 was performed
in five steps. A description of the downhole logs used and the pro-
cedures involved in each step follows, with individual steps being
listed in chronological  order.

Downhole Logs
  Temperature and resistivity variance are physical parameters that
can be indirectly related to groundwater movement. Temperature
varies with depth in the near surface as a function of the geothermal
gradient (average of 1.5T/100 ft.)2 A homogenous aquifer with
horizontal flow should show a continuous increase at this rate.
Resistivity measurements vary as a function of the total dissolved
320    SITE REMEDIATION

-------
solids (TDS) content of the groundwater. TDS is a relatively stable
component within equilibrated aquifers, generally increasing with
aquifer depth. Movement of water (either through mixing or strong
vertical gradients) can produce characteristic anomalies to the geo-
thermal gradient or TDS content in the well bore, which in turn
can be interpreted on the temperature or resistivity log as vertical
movement of groundwater in the well  bore.
  Spinner stop-counts and brine injection logs provide direct
measurements of groundwater movement. The spinner stop-count
log measures the relative vertical movement of fluid in the well
bore. The way the spinner is mounted allows movement solely from
vertical forces, with movement being expressed as a current induced
by spinner rotation. A threshold fluid velocity of approximately
0.02 ft/sec must be exceeded  to begin spinning the spinner. The
spinner stop-count log is run from the top  of the screened interval
downward. Every 20 ft. the spinner tool  is stopped and allowed
to reach equilibrium with well bore flow.  One minute of spinner
count data is collected, and the stabilized average is used as  the
stop-count for that particular depth. Stop-counts measure flow at
a particular depth without the interference of movement found in
a continuous spinner log where flow rates are measured while the
tool is pulled through the hole.
  Brine injection is performed using a tool that combines a brine
injector and a brine detector (either by resistivity or conductivity).
To estimate velocity,  a  brine slug is released a known distance
uphole from the brine detector, and the time necessary for the peak
concentration to arrive  is recorded. This method  assumes that
horizontal movement within the borehole  is insignificant and that
vertical flow velocities are quantifiable above the rates of diffu-
sion and/or vertical slippage because of  density contrasts.

Static Logs
  Geophysical logs were run under static conditions to evaluate
the groundwater movement within the San Jose 6 well bore. Four
separate logs were run, including temperature, resistivity, spinner
stop-count and brine  injection logs. For  each successive logging
run, the logging tool was decontaminated, calibrated and checked
for reproducibility and reliability. In addition, a caliper log was
run to confirm the 16-in.  diameter of the well bore.

Pump Installation
   A 125-hp submersible pump capable of a nominal pumping rate
of 2,000 gal/min was installed at a depth of 163 ft. in Well San
Jose 6. The diameter (10 in.) of the submersible system was con-
strained by the need to provide access for geophysical  logging to
the screened portion  of the well during  pumping. To facilitate
                       COLUMN PIPE FOR
                      SUBMERSIBLE PUMP
                                               4" ACCESS PIPE


                                                3/ซ" ACCESS PIPE
 DEPTH
 (IN FEET)
          912 -
SUBMERSIBLE
   PUMP
                          Figure 1
             Configuration of Well San Jose 6 After
                 Submersible Pump Installation
access, a 4-in. diameter pipe was set in the hole to a depth of
178 ft. In addition, a 3/4-in. diameter pipe was installed to a depth
of 189 ft. and equipped with a pressure transducer for water level
measurements. Figure 1 illustrates the installation of the pump and
access pipes in San Jose 6.

Dynamic Logs
  Geophysical logs were run under dynamic conditions to define
the percent water contribution of each hydrologic unit through
which San Jose 6 is screened. In addition, the pumping rate was
stepped through constant rates of 1,000, 1,500, and 1,800 gal/min
to allow  evaluation of the system under a range of hydrologic
stresses. The four logs run were temperature, resistivity, spinner
stop-count and continuous spinner. For each successive run, the
logging tool was decontaminated,  calibrated and checked for
reproducibility and reliability.
  Temperature, resistivity and spinner stop-count logs were run
and recorded three times using the same procedure as for static
logs, with the single difference of stressing the well at 1,000,1,500,
and 1,800 gal/min. Temperature and resistivity variance are still
indirectly related to flow in the well bore, while spinner stop-counts
are directly related to flow.
  A continuous spinner log was run for each of the three pumping
rates. The continuous spinner log operates on the same principles
as the spinner stop-count log. In place of discrete measurements
at routinely spaced intervals, the continuous spinner log provides
a continuous record of spinner revolutions as the instrument is
pulled up the hole at a fixed rate. Because the movement up the
borehole induces an additional velocity into the spinner, continuous
spinner and spinner stop-counts will give somewhat different values.


Depth-Specific Sampling
  Groundwater samples  obtained from Well  San Jose 6 were
analyzed for 36 VOCs and 18 inorganic parameters. Table 2 lists
the individual parameters determined in the analyses.
  Parameters analyzed were selected based  on historical  con-
taminant data from the site. Inorganic parameters were selected
to provide a comparison to other water quality types reported at
the San Jose 6 site and to support the characterization of individual
depth samples taken from the well bore.
  Suites  of geophysical logs were run on San Jose 6 during the
three pumping  periods. Spinner logs were interpreted for  each
pumping period; interpretation  allowed  for the delineation  of
contributing and noncontributing intervals. Contributing and non-
contributing intervals reflect permeable and impermeable zones,
respectively. Sample depths were chosen within impermeable zones
to obtain a discrete sample of the water quality within the bore-
hole at a specific depth. By using this procedure, a representative
sample of groundwater from each portion of the underlying con-
tributing hydrologic units was obtained.
  Water sampling was performed with a 2-1. motorized sampler.
The sampler was decontaminated at the surface and purged with
nitrogen gas. The sampler, lowered into the well by a motorized
winch to the required depth, was activated, and a sample of water
was collected. Water for the VOC samples was placed directly into
40 ml. glass vials leaving  no air bubbles.  Bottles for additional
analytical parameters were also filled directly from the sampler.
  Each sampler was decontaminated by a wash in a phosphate soap
solution, a rinse in isopropanol alcohol and finally a rinse with
deionized water.
  A QA/QC program was implemented to  provide a means of
estimating the confidence level of sample analysis results. This con-
sisted of collecting five separate types of QA/QC samples, including
sampler rinseate samples,  bottle blanks, duplicate samples,  blind
field  standards and field blanks.

Aquifer  Hydraulics
  During aquifer testing, water levels in San Jose 6  and 30
monitoring wells were measured either by hand using electronic
                                                                                                  SITE REMEDIATION    321

-------
                         Table 2
    Parameters Analyzed During Depth-Specific Sampling of
                 Municipal Well San Jose 6
                                   Inorganic Parameter!

                                 Alkalinity ICO.I  as  CaCO,
                                 Sodium (ppb)   J          *
                                 Alkalinity (HCO.J  aa CaCO.
                                 Acidity aa CaCO,  (ppm)
                                 Calcium (ppb)
                                 Nagneiium (ppb)
                                 Iron (ppb)
                                 Zinc (ppb)
                                 Potaealum (ppb)
                                 Total Organic  Carbon (ppm)
                                 Nitrate/Nitrita-N (ppm)
                                 Alkalinity aa  CaCO-  (ppm)
                                 Chloride (ppm)
                                 Suitat* (ppm)
                                 Total Diaaolved Solida  (ppn)
                                 Conductivity (micromhoa)
                                 pH (unita)
                                 Silica Diaaolv.d  (ppm)
Volatile Organic Compounds

Chloromethane
Bromomethane
Vinyl chloride
Chloroethane
Methylene chloride
Trichlorofluoromethane
Acetone
Carbon diaultide
1,1-Dichloroethene
1,1-Dichloroethane
Trana-1,2-dichloroethene
Chloroform
2-Butanone
1,2-Dichloroethane
1,1,1-Trichloroethane
Carbon tetrachloride
Vinyl acetate
Bromodichloromethane
1,2-Dichloropropene
Trana-1,3-dlchloropropene
Trichloroethene
Benzene
Dibromochloroemethane
1,1,2-Trichloroethane
Cis-1,3-dichloropropene
2-ChloroethyIvinylather
Brooofora
4-Nethyl-2-pentanone
2-Bexanone
Tetrechloroethene
1,1,2,2-Tetrachloroethane
Toluene
Chlorobenxene
Ethyl Benzene
Styrene
Total Xylenes
well sounders or automatically using pressure transducers with data
loggers. The objective of monitoring water levels during the tests
was to evaluate hydraulic connection between shallow contaminated
aquifers and the deeper aquifer used for water supply and thereby
define potential contaminant migration pathways to San Jose 6.
  Data  loggers in eight  monitoring wells were programmed  to
collect water level data at S-min. frequencies before, during and
after the test. Additionally, the pumped well and a close monitoring
well were programmed to collect data logarithmically after each
stepped increase in the pumping rate. Water levels at the remaining
21 monitoring wells were measured by hand before the test, at the
end of  the 1,000- and 1,500-gal/min steps and 2 days after an
unanticipated shutdown of the pump test. Data were not collected
at the end of the 1,800-gal/min step because the pump  failed.
Additional water level data were obtained during the  restarted
1,800-gal/min aquifer test. During the restarted test, water level
measurements  were obtained before the test, after  5 days  of
pumping and after 2 days of recovery.

RESULTS
Static Logs
  Figure 2 includes the suite of geophysical logs obtained under
static conditions.  An annotated well bore cross-section combining
interpretations made from all four static well logs is included in
Figure  3.  Interpretation of temperature and  resistivity logs
delineated the contributing/noncontributing zones as indicated by
the horizontal  arrows.  Interpretation of the brine injection log
provided the qualitative information on vertical flow and the down-
ward gradients. Only qualitative information was  obtained from
this log because of the closeness of  reported values to detection
limits. Under static conditions, the spinner stop-count log threshold
velocity (0.02 ft/sec) was not exceeded.  The average stop-count
value was zero for the entire borehole.

Dynamic Logs
   Three suites of geophysical logs were obtained under dynamic
conditions. Although minor differences  were seen between each
                                                                                              Figure 2
                                                                                       Suite or Geophysical Logs
                                                                    DEPTH IN
                                                                     Ftrr
                                                                      o	
                                                                               III


                                                                               111
                                                                                 II
                                                                                               • GAOUNOWAf ER HOW
DEPTH IN
ferr
w
no
310 •
390-
ซo •

oo
910'
MO
tn
*so
in
710
no
112

"t tt

Ml
t M
t t t
: t M
.

                                                                              STATIC           Figure 3

                                                                                       Conditions in San Jose 6
                                                                                                                  DYNAMIC
 322    SITE REMEDIATION

-------
of the three suites, general trends correlate and provide similar
results of intervals contributing water to the well and the percent
contribution of each interval. Because of the similarities, only the
logs collected at a pumping rate of 1,500 gal/min are discussed
here. Figure 2 includes the suite of geophysical logs collected at
a pumping rate of 1,500 gal/min.
  Dynamic conditions differ from static conditions in that the
reading obtained at any one depth is a function of the water volume
contributed from all aquifers downhole from that depth. An anno-
tated well bore cross section combining interpretations made from
all four dynamic well logs is provided in Figure 3. Interpretation
of the spinner stop-count log and continuous spinner log provided
the delineation of contributing/noncontributing zones as indicated
by the horizontal arrows. Under dynamic conditions, individual
temperature and total dissolved solid anomalies were not strong
enough to outweigh the cumulative dilution effects of water moving
up the  well bore. Temperature and  resistivity logs offered no
additional information.

Depth-Specific Sampling
   Seven sampling depths were selected based on the results of the
logs collected  at the three pumping rates. Depths were  selected
based on locations of noncontributing zones which divide the well
into six contributing zones. Two samples were collected within the
uppermost zone to further define water quality in the shallowest
unit contributing water to the well. Samples were obtained at depths
of 190, 230, 330,  490, 550, 660 and 730 ft.

Mass Balance Calculation
   Water quality at a specific depth in the well bore during pumping
is a mixture of the aquifer water qualities entering the well below
the sampling depth. Through mass balance calculations, it is possi-
ble to estimate the water quality in each individual unit or forma-
tion contributing to the well. Based on equations by Thomann,3
the individual formation concentration in a conservative system
will be  equal to:
  where:
                         Qf.n
  [Fn]  = formation concentration in the nth layer
  [WJ = well bore concentration at the nth layer
  QTn = total flow in well bore at nth layer  depth
  Qf n  =  individual layer flow from nth layer

  Since well bore contaminant concentrations were obtained during
depth-specific sampling, only flow measurement data were needed
to provide the estimate of formation concentration. Continuous
spinner log data were used to provide values for flow at every point
downhole. Percent flow was interpreted on each of the three spinner
logs. The cumulative value for percent flow at each of the seven
sampling  depths, in  addition  to individual contributions,  was
obtained. To minimize any errors possibly unique to an individual
logging run, the three depth-specific cumulative flow values were
averaged  and are given in Table 3.
                         Table 3
          Average Percent Flow From Spinner Logs
        Depth
     Interval
       (feet)

       190-230
       230-330
       330-490
       490-550
       550-660
       660-730
       730-812
Percent
 Flow

  9.6
 19.7
 26.5
 12.9
 16.4
  9.9
  5.1
                                           Cumulative
                                          Percent Flow
                                             100.1
                                              90.5
                                              70.8
                                              44.3
                                              31.4
                                              15.0
                                                5.1
                                                                   Aquifer Hydraulics
                                                                      The two graphs in Figure 6 illustrate head differences between
                                                                   paired wells under static and pumping conditions. These well pairs
                                                                   show lower water levels in the deeper wells, indicating downward
                                                                   gradients. In addition, the graphs show an increase  in the head
                                                                   difference between wells of approximately two to three times after
                                                                   4  days of pumping.
                                                                      To obtain flow rate, percent flow values were multiplied by 1,800
                                                                   gal/min, which was the pumping rate during sampling.

                                                                   Organic Results
                                                                      Mass balances were calculated (Eq. 1) for the four detected vola-
                                                                   tile organic constituents.  Table 4 summarizes the results.
                                                                                               Table 4
                                                                         Formation Concentrations From Mass Balance Calculations
                                                                                                  Lionซ I
                                                                           conซtituซnt
                                                                     l,l-DiehloriMthซM O.1-DCE) Ippb)
                                                                     1,1-DlchlorMthiiM 11,1-DCA) (ppbt
                                                                     Triehlorocthora (TCt) Ippb)
                                                                     Tซtrซchloreซthซn* (FCK) Ippb)
                                                                     Sodiun IppHl
                                                                     Blcarbonit* (pp>)
                                                                     Calelui Ippel
                                                                        iui (ppa)
                                                                     line lpp.1
                                                                     rotauitui Ipiw)
                                                                     Hitro9ซn-NO,/NO. lppป)
                                                                     Chloride (ppa) '
                                                                     Sulfซtซ (p|Ml
                                                                     silica (PPM)
                                                                     TDI lpp.1
                                                                     spcrlflc Conductivity 
-------
probable explanation for this is that contamination in  the five
intervals originated from the same source, probably migrating down
the well bore from the shallower to the deeper contributing units.

Inorganic Results
  Mass balances were  calculated  (Eq.  1) for  13  inorganic
parameters. Table 4 summarizes the results.
  Two standard geochemical equilibrium calculations were used
as a check on the validity of the inorganic formation concentra-
tions: ionic imbalances, and  a Piper trilinear diagram.4 Ionic
imbalance errors ranged from 1.2 to  14.0%.  The Piper  trilinear
diagram (Figure 5) indicates the calcium/sulfate nature of water
at the San Jose 6 Superfund site.
                        CLUSTERED WELL PAIR ป1
               — cซ
                  •OKIMTAK NEACTDM vtLUU
                          Figure 5
    Piper Trilinear Diagram San Jose 6 Formation Concentrations
  To interpret the distribution of contaminants in San Jose 6, the
similarity between the three shallowest samples and the two deepest
must be taken into account. The Piper trilinear diagram shows a
pattern that  suggests a diluted relationship.  Depths 190 and
730 ft. cluster fairly well, while depths 230, 330 and 660 ft. take
up various positions along a mixing path caused by dilution with
waters from 490 and 550 ft. These data suggest that shallow and
deep waters share a similar source. Differences between the two
arise because of dilution encountered along the migration pathway.

Quality Assurance/Quality Control
  For VOCs, rinseate and blank samples indicated no contamina-
tion with indicator compounds (1,1-DCE; 1,1-DCA; TCE; PCE).
Duplicate sample replication showed close correlation, and blind
field standard recovery was satisfactory. Inorganic QA/QC samples
yielded results with as high or a higher level of confidence. In short,
QA/QC  data  support both magnitude and occurrence of each
identified constituent used to interpret  chemical analyses.
  All graphs in Figure 6 show drawdowns in monitoring wells in
response to San Jose 6 pumping. The hydraulic responses, coupled
with the vertical gradients, indicate that groundwater in the con-
taminated overlying aquifer leaks vertically into the intervals from
which San Jose 6 produced water. In addition, the data show that
the rate of leakage increases when Well San Jose 6 is pumped.
  This information demonstrates that contaminants in the over-
lying aquifer have a flow path to the first screened interval of San


i
T.
gf 4ป
g| ซ•ป•
i 9 * *"
ซ —
<
I







"^









^-









•*ป•,



















<—









^









"N^






^


.
\








c

\



MM
V


CM


\


OIA
•V


wei



^

rcw


j.




^
tu.
/




,
/
'


•• n|





f~*



^~




f




_x




>




*s.



1*^





/N








                                                                                       I    1    I    1
                                                                                              (TMOUtANOSI
                                                                                        riuc SMCI TOT STAJTOD
                                                                                          CLUSTERED WELL PAIR ri















H2 •

"S_















*^N
^^















•— *""
_ป















^-
"•^














r-














^r
~s















^















\
\\
I
\
\













^
\

\
\
\












\
\


\
\
^











\
\


\
\
\











^
s



K








/
/
/
f ,
/
/
/
i
•









p*i











/


^
r










/


s












\


>












t\











                                                                                              ITMOUSANOSI
                                                                                       T.MC SJMCf TtST STAATED IHINUTtS)
                                                                                             Figure 6
                                                                          Clustered Well Pair Response (o San Jose 6 Pumping
Jose 6 under static conditions and that rate of flow along this path
is enhanced when San Jose 6 is pumped.

SUMMARY OF RESULTS
  Data collected from the San Jose 6 aquifer test in April 1987
provided additional information regarding the nature, extent and
migration of groundwater contaminants at the San Jose 6 Super-
fund site. The following significant findings  were generated from
the aquifer test:

• Water produced from San Jose 6 after 5 days of pumping con-
  tained total VOCs of approximately 18 pg/1
• VOCs detected in depth-specific samples collected from San Jose
  6 after 5 days of pumping included:

    - Tetrachoroethene (PCE)
      Trichloroethene (TCE)
      1,1-Dichloroethene (1,1-DCE)
    - 1,1-Dichloroethane (1,1-DCA)

• The vertical extent of VOCs in the formation adjacent to San
  Jose 6 was from 180 to 490 ft. and from  660 to 730 ft. below
  grade. No VOCs were detected in the interval from 490 to 660
  ft. below grade.
324    SITE REMEDIATION

-------
• Groundwater in the contaminated overlying aquifer leaks verti-
  cally downward into the intervals  from which San  Jose  6
  produces water. This leakage occurs under static conditions when
  San Jose 6 is not pumping, and it increases when San Jose 6
  is pumped.
• Under static conditions, groundwater flows vertically from con-
  taminated hydrologic unit of the higher hydraulic head through
  the screened well bore of San Jose 6 to underlying hydrologic
  units of the lower head. This allows contaminants to migrate
  vertically in the San Jose 6 well bore and accounts for some of
  the deep contamination detected in San Jose 6.

CONCLUSION
  Data collected  during  the  April 1987 aquifer  test  provided
valuable information regarding the vertical nature, extent and
migration  of contamination at  the San  Jose 6  Superfund site.
Specifically, temperature, resistivity, spinner stop-count and brine
injection logs defined the water movement through the well bore
under static and pumping  conditions.  Depth-specific  samples,
coupled with the flow results, defined the contaminants present
in the six major hydrologic units from which San Jose 6 produces
water.
  This approach provided a rapid and cost-effective characteriza-
tion of groundwater contamination in Well San Jose 6. The more
conventional approach of installing the sampling wells in  each
hydrologic unit would have been more costly and time-consuming.
  In addition, monitoring of water levels in surrounding observa-
tion wells during the aquifer test aided in defining the mechanism
by which contaminants reached the producing aquifers in San Jose
6. Results from the pump test, coupled with other components of
the remedial investigation, will provide the needed information to
evaluate and design technically feasible and cost-effective remedial
actions for the San Jose 6 Superfund site.
ACKNOWLEDGEMENTS
  The authors would like to acknowledge the help provided by
U.S. EPA Region VI staff, WELENCO Logging Co. and Layne-
Western Co. Each provided additional insight and logistical help
in accomplishing  the aquifer test.

References
1. Kelley, V.C. "Geology of Albuquerque Basin, New Mexico." New
   Mexico Institute of Mining and Technology, Memoir 33. 1977.
2. Freeze, R.A. and Cherry, J.A., Groundwater. Prentice-Hall, Inc., Engle-
  wood Cliffs, NJ, 1979.
3. Thomann, R.V.  System Analysis and Water Quality Management.
  McGraw-Hill Book Co., NY, 1983, 61-69.
4. Piper, A.M. A Graphic "Procedure in the Geochemical Interpretation
  of Water Analyses," Trans. Amer. Geophys.  Union, 1944, 914-923.
                                                                                                 SITE REMEDIATION    325

-------
                                           A  Phased Approach
                                    To Remedial  Investigations:
                                  Focusing Effort  and  Reducing
                             Overall  remedial  Investigation  Costs

                                                  Gary H off master
                                   Michigan Department of  Natural Resources
                                          Groundwater Quality Division
                                                 Lansing, Michigan
                                                Leonard C.  Johnson
                                                 Jane M. Patarcity
                                                 Robert J. Hubbard
                                                 NUS Corporation
                                       Waste Management Services Group
                                             Pittsburgh,  Pennsylvania
  ABSTRACT
   The  Michigan Department of Natural Resources and NUS
  Corporation successfully implemented a "phased"  Remedial
  Investigation at an abandoned hazardous waste site in Monroe
  County, Michigan. This phased approach significantly reduced
  costs and reduced the time needed to implement the study through
  savings in sample analysis cost and monitoring well installation.
  Surface geophysical surveys including resistivity, seismic refrac-
  tion, magnetics and electromagnetic conductivity profiling were
  conducted to characterize buried wastes, geologic and hydrogeo-
  logic conditions.  Shallow  groundwater,  subsurface  soil and
  monitoring well samples were screened in the field for volatile
  organics using a gas chromatograph (GC). The screening data were
  used to determine drilling locations and the extent of contamina-
  tion in environmental media. Surface water, sediment and test pit
  samples were submitted  for  complete  laboratory  analyses.
  Monitoring well samples were submitted for laboratory analysis
  to confirm field analysis results.
   Quantitative risk assessment techniques were used to evaluate
  the Phase I data. The results indicated little environmental impact
 and low public health risks as a result of contaminant migration
  from the site. These results enabled a significant decrease in the
 scope of the investigation and resulted in a cost savings of approxi-
 mately 40<7o compared to the original estimate.

 INTRODUCTION
   Recognition of the adverse environmental impact of common
 waste disposal practices from  the late  1960s  through the 1970s
 fostered concerted  efforts in the 1980s to identify and remediate
 sites where the public health  and the environment are threatened
 by uncontrolled hazardous wastes. In October  1982, the Michigan
 State Legislature passed Act No. 307 of the Public  Act of 1982,
 known as the Michigan Environmental Response Act. This act
 provided for the identification, assessment and prioritization of
 sites of environmental contamination within the State  of Michigan.
 The Act also provided for response activity, where  appropriate,
 at these sites. Under the direction of that bill, the State of Michigan
 developed a model  that assessed the relative hazards at each site
 and assigned a numerical score to it. The sites were ranked and
 placed into two groups—one for remedial investigation, and the
 other for remedial action. After public review,  866 sites were
 included on the two lists.
   Some of the sites on the Michigan Priority List are also on the
 NPL and are eligible  for funding under both programs The
 remaining Michigan Priority List sites are only eligible for funding

326   SITE REMEDIATION
under the Michigan program. The Stevens Landfill Site Remedial
Investigation was conducted under the Michigan Act No. 307
funding program.
  NUS Corporation, as a prime contractor to the Michigan Depart-
ment of Natural Resources (MDNR) under the Act 307 Program,
developed a phased investigation program  for the Stevens Land-
fill Site. The approach used for the Stevens Landfill RJ is the same
one being used by the U.S. EPA  on many RI/FS projects in
response to the SARA.
                        Figure 1
                General Arrangement of the
                 Site and Surrounding Area

-------
DESCRIPTION OF THE SITE
  The Stevens Landfill Site is situated in Monroe County, Michi-
gan, about 2 miles north of Toledo, Ohio. The site is a 34-acre
unlicensed dump that operated as a co-disposal  facility (i.e.,
municipal and industrial wastes) between  1957 and 1965. The
immediate area surrounding the site is relatively flat, with eleva-
tions ranging from 625 to 640 ft. above mean sea level. A marshy
area of approximately 60 acres lies immediately east of the site.
An intermittent stream, Salter Drain, crosses the southern portion
of the site. Salter Drain and a tributary drain the  landfill area,
flowing  east into Indian Creek approximately 1.5 miles down-
stream. Drainage in this area is poor, with indication of possible
inundation and/or a high water table. A dense growth of brush
and trees exists on the landfill. Overturned trees show domestic
refuse lodged in their root systems.
  MDNR first visited the site in February 1980. Approximately
1,100 drums were observed in two distinct areas near the southern
and northern borders of the  landfill. Also, numerous buried pits
were suspected to exist and to have received barrels, liquid chemi-
cal wastes and even discharge from tank trucks. Figure 1 shows
the general arrangement of the site and surrounding area.
  Because many of the drums had rusted through  or been
punctured, their contents apparently had leaked onto the ground.
While the majority of the drums were on the surface of the land-
fill, some were partially buried or in standing water. Local residents
also reported that the landfill accepted liquid wastes that were
flushed directly into the landfill. These alleged and apparent releases
were believed to have  resulted  in contamination of surface soils,
subsurface soils and groundwater which could present hazards to
the public health and environment through various migration routes
and exposure pathways.
  During the period from 1980 to 1982, several sampling efforts
by Federal, state and local environmental and public health agencies
resulted in the collection of both environmental and drum waste
samples. The results of these efforts indicated that many of the
chemicals identified in the drums were also found in the local soils,
sediments and groundwater. However, the data did not adequately
characterize the relationship between the wastes found on site and
the contaminants identified in  the environmental media. MDNR
completed a drum-removal action concurrently with the RI in which
all the surface and partially  exposed drums were removed from
the site.
TECHNICAL APPROACH AND RATIONALE
FOR THE REMEDIAL INVESTIGATION
  The RI objectives were developed based upon existing informa-
tion on  past disposal practices, the suspected nature of the con-
taminants, the apparent areal distribution of the drums and existing
information on regional and site-specific hydrogeologic conditions.
The primary objectives of  the Stevens Landfill Site  Remedial
Investigation, as listed in the October 1985  Work  Plan, were:

• Delineate approximate areal extent of the landfill
• Determine (approximately) the continuity, extent, depth, thick-
  ness and permeability of a clay layer reported to exist beneath
  and in the vicinity of the  landfill
• Hydraulically characterize the upper (sand) aquifer beneath and
  in the near vicinity of the  landfill (depth to water table,  water
  table gradient, aquifer thickness, permeability, flow direction,
  flow velocity
• Determine the nature and extent of site-related chemical con-
  tamination within the upper aquifer
• Determine the potential for the upper aquifer to be hydraulically
  connected to the lower (bedrock) aquifer
• Determine the approximate nature and extent  of site-related
  chemical contamination in surface water (and sediments) that
  traverse the site and/or are contiguous with the site
• Compile a data base that will support a site-related Public Health
  and Risk Assessment.
• Conduct a site-related Public Health and Risk Assessment that
  would evaluate the following:

  - nature and extent of contaminant source's
  - existing and potential pathways for contaminant transport
  - location of existing and potential receptors
  - risk to receptors associated with existing or potential exposure
    to site-related contaminants

• Compile a data base that will support a Feasibility Study of
  candidate remedial alternatives

  As a result  of the anticipated effort required to evaluate the
complex bedrock aquifer, the Phase I RI focused on evaluation
of contaminant transport within the upper sand aquifer. Based on
the Phase I RI results,  the need to conduct additional work to
characterize the lower bedrock aquifer was assessed. This poten-
tial additional work represented the Phase II RI effort. Phase III
work was also planned to further characterize the hydraulic con-
nection between the overburden and bedrock aquifers and included
exploratory test pitting to further delineate subsurface conditions
at the site.
  The RI  activities, as proposed in the October  1985 Work Plan,
included the following:


• Survey landfill property boundaries, establish a 50-ft. grid for
  geophysical  surveys  and sampling  operations and establish
  ground control for aerial photography and topographic mapping
• Conduct geophysical surveys to aid in the characterization of
  buried  wastes  and hydrogeologic conditions. The  primary
  objectives of the survey were:  the definition of the areal extent
  of buried refuse; the delineation of areas where large amounts
  of ferromagnetic materials were buried (indicating possible lo-
  cations of the suspected buried drums and tanker trucks on site);
  detection and definition of conductive plumes of contaminated
  groundwater and the determination of the depth, thickness and
  continuity of the clay overlying the limestone bedrock. The sur-
  face geophysical results were to be used in conjunction with a
  soil-gas  testing program to aid in the determination of monitor-
  ing well locations and to determine the need for conducting ex-
  ploratory test  pits.
• Conduct soil-gas testing to aid in the location and delineation
  of the potential contamination source and groundwater plume.
  The data also were to be used as an aid to locating the proposed
  groundwater monitoring wells

  Drill and install 18 monitoring wells  in the sand unit overlying
the bedrock. These wells were to be used to obtain data on the
location of contaminant sources, to determine the continuity and
permeability of the clay layer and  to  determine chemical and
hydraulic  characteristics of the sand unit.
• Conduct two rounds of sampling at  the following locations:

  - Five on-site ponds
  - Two  locations in Salter Drain downstream of the site
  - One location in Salter Drain upstream of the  site
  - Three existing monitoring wells
  - Eighteen Phase I shallow monitoring  wells
    Two  locations in the flooded quarry south of the site

• Evaluate data from Phase I activities  to determine the need for
  Phase II activities
• Conduct Phase II drilling and sampling in the bedrock aquifer
  and off-site surficial aquifer
• Conduct Phase III pumping tests to determine the degree of
  hydraulic connection between the two aquifers
• Conduct Phase III exploratory test pitting to further determine
  the subsurface conditions at the site in relation to contaminant
  sources
                                                                                                 SITE REMEDIATION    327

-------
REMEDIAL INVESTIGATION SUMMARY
  The RI for the Stevens Landfill Site was conducted jointly by
MDNR and NUS. The MDNR Geotechnical Group assumed the
lead in the seismic refraction survey. MDNR also performed a
portion  of the environmental sampling and analysis.  MDNR,
through several additional contractors, performed drum-removal
operations,  air sampling and analysis, and soil sampling and
analysis.
  NUS  and its subcontractor  performed  the  hydrogeologic
investigation at the site,  which  included drilling, installing and
sampling groundwater monitoring wells. NUS was responsible for
the remaining work, which included environmental sampling and
analysis, geophysical surveys (resistivity, magnetometer and con-
ductivity), test pit excavation and sample analysis, risk assessment
and report preparation.  The following activities were conducted
at the site in accordance with the technical approach outlined above,
and to accomplish the drum-removal action.

• October 1985: NUS/MDNR collected five surface water samples,
  one groundwater sample and 11 sediment samples from locations
  within and adjacent to the Stevens Landfill Site. Samples were
  analyzed  for U.S. EPA  priority  pollutants by  the NUS
  laboratory.
• November-December  1985:  MDNR removed approximately
  1,100 barrels from the site and performed compatibility tests on
  approximately 462 drum waste samples. Three rounds of air
  samples  were collected from various locations  (upwind and
  downwind) before and during drum staging and during drum
  loading. Twenty surface soil samples from former drum disposal
  areas also were collected.
• December 1985: NUS performed a surface resistivity survey
  consisting of six vertical electrical soundings (VES) along the
  western edge of the site.
• December 1985:  MDNR  collected  16 drum  waste  samples.
  Samples were analyzed for U.S. EPA priority pollutants by the
  NUS  laboratory.
• December 1985:  NUS conducted  a  shallow  groundwater
  screening program that  involved collecting  35 groundwater
  samples on a 200-ft. grid and screening each sample for selected
  volatile organic compounds. This program replaced the soil-gas
  sampling program due to high water table conditions at the site.
• January  1986:  MDNR collected  six surface  soil  samples  at
  previous  sample locations. Samples were analyzed for EP toxicity
  inorganics.
• January  1986: MDNR performed a seismic refraction survey.
• February 1986: MDNR collected two surface soil samples from
  the former drum-staging area. Samples were analyzed  for EP
  toxicity inorganics.
• April 1986: NUS conducted a magnetics survey on a 50-ft. grid
  and acquired approximately 1,500 total magnetic-field and 1,500
  magnetic-gradient measurements. NUS also  conducted
  electromagnetic-conductivity  profiling and acquired approxi-
  mately 500 measurements in  both the horizontal and vertical
  dipole configuration along survey lines across the site.
• May  1986: MDNR collected six surface soil samples. Samples
  were subjected to Michigan scans 1,2 and 3 and various inorganic
  analyses.
• June  1986: NUS/MDNR performed a Phase I drilling program
  that resulted in seven bore  holes and  10 monitoring wells.
  Gamma-ray logging and field  GC screening of groundwater for
  a wide range of organics were  conducted in each drill hole. Clay
  samples were collected in each drill hole using Shelby tubes, and
  each  sample was analyzed for various chemical and physical
  properties.
 •  July  1986: MDNR collected  10 surface soil samples. Samples
  were subjected to Michigan scans 1,2 and 3 and various inorganic
   analyses.
 •  July  1986: NUS collected nine groundwater samples from the
  nine recently installed monitoring wells. Samples were analyzed

 328    SITE  REMEDIATION
  for U.S.  EPA priority pollutants by the NUS laboratory.
• July 1986: NUS collected 10 groundwater samples from the new
  monitoring wells and an existing monitoring well on the site.
  Samples were analyzed for U.S. EPA priority pollutants by the
  NUS laboratory.
• October  1986: MDNR collected seven surface soil samples.
  Samples were subjected to Michigan scans 1, 2 and 3.
• January 1987: NUS/MDNR installed 10 test pits  at the locations
  of magnetic anomalies identified during the geophysical surveys.
  Three soil samples and two liquid waste samples (drum wastes)
  were shipped to the NUS laboratory for U.S. EPA priority
  pollutant analysis.

  Field and laboratory data were evaluated as they became avail-
able. This information was used to guide subsequent activities.
Activities specified in the technical approach were modified or
deleted accordingly.
  As can be seen by reviewing the activities, no Phase II activities
and one Phase III  activity were performed.

RESULTS OF THE REMEDIAL INVESTIGATION
  Phase I of the RI  revealed the following information with regard
to the site. Site-specific geology consists of lacustrine deposits of
Pliestocene age overlying limestone bedrock of the Silurian age Bass
Islands Group. The lacustrine deposits vary from  30 to 45 ft. in
thickness in the vicinity of the landfill. The Bass  Islands Group
is a regional aquifer used for domestic water supply. Three primary
geologic units were encountered  in  the lacustrine deposits: a
"surface sand," a  "silty sand" and  a "silty  clay."  There is a
gradual reduction in grain size from top to bottom in the lacus-
trine deposits at  the landfill.
  Shelby tube samples form the lowermost silty clay unit yielded
permeability  results  ranging from 4.01 xlO~8  to   1.2xlO"8
cm/sec. Thus, the stratographic  framework  of  the  lacustrine
deposits was found to be favorable to the exclusion of contaminants
from the bedrock regional aquifer.
  Groundwater in  the upper sand units within  the  lacustrine
deposits was found to be unconfined.  Groundwater flow in these
units is to the south at an estimated velocity ranging from 0.004
to 0.009 ft/day. In-situ permeability measurements ranged from
8.8xlO6 to 3.45 x 10-4 cm/sec.
  Drum waste samples contained high concentrations of toluene,
ethylbenzene, methylene chloride, xylene, s'tyrene, naphthalene,
bis(2-ethylhexyl)phthalate, lead, chromium and zinc.
  Minimal contamination by organic and inorganic chemicals was
detected in groundwater at the landfill in the lacustrine deposits.
Migration of chemical contaminants from the lacustrine deposits
to the regional bedrock aquifer is unlikely based on the low con-
centrations and the impermeable nature of the lowermost silty clay
unit.
  Surface water and sediment samples collected on-site and from
Salter  Drain off-site did not indicate site-related contaminant
migration and/or accumulation in these  media.
  Surface soil in the drum  disposal areas was contaminated by
volatile organics, pesticides, PCBs and metals.
  Air  sampling and  analysis before  and during  drum-removal
operations indicated that the drums were a source of contamination
of volatile organics. However, the drums have since been removed
and  no further releases are likely.
  As a result of the Phase 1 investigation, it became apparent that
contamination at the site had not impacted the groundwater in the
upper water-bearing zone to an appreciable degree. Furthermore,
the confining (clay) layer was found to be relatively impermeable
and continuous across the site. Therefore,  it was  concluded that
the Phase II and Phase III  activities to characterize the bedrock
aquifer and its relationship  to the upper water-bearing zone were
unnecessary. Results  of the Phase III test-pitting  operation indi-
cated that only one of the 10 test pits contained waste materials.
Sample analysis results indicate materials similar to that found in

-------
the surface drums.

QUANTITATIVE RISK ASSESSMENT
  Risks to human and environmental receptors posed by chemical
contaminants at or originating from the Stevens Landfill Site were
assessed by considering three major aspects of the contamination
and the environmental fate and transport of site chemicals:

• The potential for and concentration to which receptors may be
  exposed
• The carcinogenic and noncarcinogenic human health hazards and
  potential environmental impacts of site chemicals
• The risks associated with exposure as compared to applicable
  regulatory standards and guidelines for the protection of human
  or environmental receptors
• The assessment approach used is the one suggested by the U.S.
  EPA and consists of hazard identification, dose-response evalu-
  ation, exposure assessment and risk characterization.

  Risks associated with the presence of carcinogenic and noncar-
cinogenic chemicals are well represented by the Hazard Indices and
quantitative carcinogenic risk estimates  developed based on the
Phase IRI data. The incremental cancer risk associated with dermal
contact with contaminated surface soils at the site was estimated
at 2x 1O~7. Hazard Indices for dermal contact with soils  are all
at least one order of magnitude less than unity. The Hazard Indices
provide an estimate of noncarcinogenic risk through comparison
with the Acceptable Daily Intake. The low values of the Hazard
Indices  indicate that noncarcinogenic  chemical  doses will not
approach the ADIs.
  The risks posed if the surficial sand unit were developed for
potable use primarily arise from the presence of lead and nickel.
The Hazard Indices for both of these chemicals exceed unity for
the exposure scenario considered. However, there is no apparent
migration pathway from source to receptor under existing site con-
ditions since the sand unit is not used as a potable water  source
and the very low permeability clay prevents vertical migration to
the bedrock regional aquifer. Thus, it is apparent that these noncar-
cinogenic risks represent future potential risks as opposed to current
risks.
  Of the hazardous organic chemicals detected in groundwater,
only bis(2-ethylhexyl)phthalate was detected in monitoring well
samples. The remaining chemicals were detected only in samples
subjected to field  screening.  Because of the absence of mass-
spectrometric confirmation of the presence of these chemicals in
the field-screened samples, their presence in site groundwater is
somewhat suspect. This fact notwithstanding, the carcinogenic risk
estimates for ingestion of groundwater ranged from 5 x 10~6 to
3xlO-8.

CONCLUSIONS

  An important aspect of the Phase I investigation was the use
of inexpensive and innovative data collection techniques such as
surface geophysics, shallow  groundwater sampling,  and field
screening of the shallow groundwater samples and subsurface soil
samples from  monitoring well  soil borings  using a mobile gas
chromatography laboratory. Data from these activities as well as
data from the first round of sampling were used to guide subsequent
Phase I data-collection activities within the Phase I investigation,
which resulted in the elimination of 10 Phase I shallow monitoring
wells and a second round of surface water and sediment sampling
and analysis.
  Evaluation of the extent and magnitude of contamination in
environmental media, based on the RI data, indicated that con-
tamination at the Stevens Landfill Site is of very low concentrations,
has not migrated off site to any appreciable extent and yields very
low risk estimates. The low site-related contamination detected in
groundwater and the existence of a continuous, thick,  very low-
permeability clay layer separating the low levels of groundwater
contamination in the  surficial sand from the regional bedrock
aquifer provided sufficient basis to eliminate Phase II drilling and
sampling and Phase III aquifer pumping tests from the RI program.
  The end  result of this approach was a significant reduction in
the overall  cost of the RI when compared with the original esti-
mate.The savings for the Stevens  Landfill Site RI  has  been esti-
mated at 40%.
REFERENCES

1. NUS Corporation. "Work Plan, Remedial Investigation/Feasibility
   Study. Stevens Landfill Site, Monroe County, Michigan" NUS Report
   No. R-33-6-5-1, Pittsburgh, PA, 1985.
2. NUS Corporation, 1987. "Draft Remedial Investigation Report, Stevens
   Landfill Site, Monroe County, Michigan" NUS Report No. D-33-2-7-1,
   Pittsburgh, PA,  1987.
                                                                                                  SITE REMEDIATION    329

-------
                        Verona  Well  Field—Where  the  Action Is

                                                  John C. Tanaka
                                     U.S. Environmental  Protection Agency
                                                  Chicago,  Illinois
                                              Joan van  Munster, P.E.
                                                     CH2M  Hill
                                                  Denver,  Colorado
                                                 Alan Amoth, P.E.
                                                     CH2M  Hill
                                                  Corvallis,  Oregon
ABSTRACT
  The Superfund program is rapidly entering the era of remedial
action. Although its progress has been maligned, many projects
are approaching the remedial design and construction phase. Avail-
able practical experience and guidance now include techniques for
remedial investigations, calculations for public health assessments
and feasibility studies. However, similar guidance does not exist
for remedial action projects.
  This paper discusses the design and construction of a ground-
water extraction system, a source  control operable unit project
successfully completed under the Superfund program. However,
the following discussion is more than just another case history;
the paper provides practical knowledge and specific lessons learned
while preparing the remedial design and implementing the remedial
action.
INTRODUCTION
  The Verona Well Field project in Michigan is unique since it is
the first REM IV project for which the engineer (CH2M  HILL)
has prepared the remedial design and contract bid documents (plans
and specifications) and been responsible for resident construction
management services. During the remedial design, elements unique
to a hazardous waste project (such as health and safety require-
ments) had to be incorporated into the contract documents. The
bid period provided valuable information regarding the willing-
ness and ability of construction and hazardous waste contractors
to perform remedial actions under Superfund. During the construc-
tion  phase, the construction management  activities  had to be
tailored to hazardous waste  site work. This paper discusses the
aspects of the design, bidding and construction phase activities that
are unique to a Superfund project.

SITE HISTORY
  The Verona Well Field provides potable groundwater to approxi-
mately 35,000  residents and several major businesses in Battle
Creek, Michigan. In 1981, during routine testing of the city's water,
volatile organic chemical (VOC) contamination was discovered.
   In the summer of 1984, an initial remedial measure (IRM) was
con-structed at the existing Verona wells. The purpose of the  IRM
was to prevent an advancing VOC plume from contaminating more
Verona wells. The IRM included construction of an airstripping
system with vapor phase activated carbon, to treat extracted
groundwater before discharge to the Battle Creek River. The system
was designed to handle a maximum flow of approximately 2,500
gal/min.
   In June 1985, an operable unit feasibility study, which examined
potential source control measures at one of the major sources of
well field contamination, was published. This VOC source was a
facility that stored, transferred and packaged a variety of chlori-
nated and  nonchlorinated  solvents.  Twenty-one underground
storage tanks were used for bulk storage. Total VOC contamina-
tion in groundwater was found in concentration as high as 100mg/l
at the facility.
                                         VEftONAWeUnCLD
                                         surammo STTE
                          Figure I
 330    SITE REMEDIATION

-------
  The selected remedial action alternative (August 1985 ROD) was
a combined groundwater extraction (OWE) and  a soil vapor
extraction (SVE) system. Construction of the OWE system (which
began in October 1986) consists of nine extraction wells, including
one large, fully  screened product recovery well. The system is
designed  so that  the extracted groundwater can be treated by the
existing air stripper. Five wells are located on the property of the
source facility, and four wells are located off-site. Extracted water
is transported about 1 mile through a high-density polyethylene
(HOPE)  pipe to the existing air stripper.
  The initial high concentrations of VOC's in extracted ground-
water could not  be handled by the air stripper without pretreat-
ment. Therefore, a  liquid  phase activated carbon system was
installed  at the air stripper to treat highly contaminated ground-
water prior to final treatment.
REMEDIAL  DESIGN
  The first phase of implementing the selected remedial alterna-
tive was  to perform the remedial design. The design work was
divided into two  packages: the OWE system and the SVE  system.
  While preparing the contract documents for the GWE  system,
the project team noticed some significant differences in these docu-
ments as compared  to non-Superfund site design work. Some
differences occurred because part of the work was to be conducted
on a hazardous waste site, while other differences occurred  because
this was an EPA project. The following points summarize  these
differences:

• The front-end documents were more comprehensive than for a
  traditional non-hazardous waste project for two reasons: first,
  the project  is U.S. EPA-sponsored  and many  requirements
  incorporated into the general conditions were adapted from the
  Federal Acquisition Regulations and other government require-
  ments;  and second,  the project  work  involved potentially
  hazardous conditions, necessitating many extra requirements.

  The general requirements included a detailed discussion of health
and safety procedures. In the Verona  contract documents,  the
health and safety requirements  included providing an air moni-
toring program,  a site safety officer, training and  physical certifi-
cations, active work area monitoring and daily site safety reporting.
  Selection of transfer piping that contained adequate flexibility,
strength,  chemical resistance and leak resistance was made. HDPE
piping was used  due to the VOCs in the process stream. .
  A portion of the groundwater carrier pipe was installed  in an
existing storm sewer because of railroad right-of-way issues and
because the state required the pipe to be double-lined. Part of the
design effort was to investigate the stormwater pipe to see if it could
accommodate the new pipeline.
  Coordination with the state, county and city regulatory person-
nel, in addition  to the U.S. EPA, was  essential on this project.
It involved traffic planning; establishing air monitoring require-
ments; determining what process variables needed to be monitored,
controlled, and  alarmed;  and identifying instrumentation and
control equipment.
  After  preparing  the  plans and specifications,  the engineer
prepared  a detailed capital cost estimate of construction, including
adjustments for work to be performed under hazardous conditions.
The final engineer's  estimate  was $1.4  million.
  Construction management was planned at the same time that
the plans  and specifications were prepared. The scope of construc-
tion management activities was incorporated into the "General
Conditions." On Superfund projects, the responsibilities of the
construction manager are more comprehensive than on a traditional
project,  since the U.S. EPA views the construction manager
essentially as  the "general contractor."

BIDDING PROCESS
  The bidding process on this project began in mid-May 1986 with
an advertisement in the Commerce Business Daily requesting state-
ments of qualifications (SOQs) from firms interested in construc-
tion work at a hazardous waste site. Bid documents were sent to
11 firms that had been prequalified based on their SOQs. Two bids
were received after a 3-week bid period, and both bids were from
construction-oriented firms as opposed to hazardous waste cleanup
firms. The bids were rejected because the contractors could not
comply with the proposed schedule, and the bids were determined
to be nonresponsive. The remaining nine firms, which were pri-
marily cleanup firms, were contacted to find out why they declined
to bid. The primary reasons they cited for not being interested in
bidding  were as  follows:


•  More than 50% of the general construction work was not on
   the hazardous waste site, and the off-site work did not interest
   the cleanup firms.
•  Many of the firms were busy with other work or had cleanup
   projects on the horizon.
•  Some firms were concerned with the liability associated with the
   work  and did not want to rely on the government's third party
   indemnification; in addition, insurance and bonds in the required
   amounts were, in some cases, difficult to obtain at a reasonable
   cost.
•  Since  most of the work would have to be subcontracted out, the
   firms  were concerned about obtaining reliable sub-bids, the risk
   involved with managing a large amount of subcontracted work
   and the problem of finding construction subcontractors that were
   health) and safety-trained to perform the on-site portions of the
   work.
•  There were extensive  requirements in the contract documents
   related to the health and safety procedures to be implemented
   on-site, which the less-experienced firms  figured would be too
   costly to implement.
•  The cleanup firms felt that because of the large amount of con-
   struction work involved, they would be competing with many
   local  contractors and  they could not be cost competitive.

   The project was rebid in August 1986 to six of the original
11 firms,  but only the same two bidders responded. Relatively
minor changes were made to the contract documents. The low bid,
which was from Loughney Dewatering, Inc. of  Baldwin, New
York, was accepted for $1.827 million. Although the bid was 30%
over the engineer's estimate, a contract was awarded since a con-
certed effort had been made to obtain competitive bids.
   After  the project was completed, the construction contractor was
asked to explain its approach to bidding the project. The following
information was provided:

•  The worst-case conditions were assumed for the hazardous waste
   portion of the project in terms of crew size, the level of protec-
   tion required and the time required to complete the work under
   hazardous conditions.
•  The safety equipment, monitoring equipment and decontami-
   nation facilities (such  as the concrete pad for decontaminating
   equipment and shower  facilities) may cost $50,000 or  more,
   depending on how they are specified in the contract documents.
•  The medical monitoring requirements, training requirements, on-
   site safety officer  and similar  requirements can  run between
   $75,000 to $100,000 for a medium-size project. (Note: This was
   before passage of 29 CFR ง1910.120, which requires a minimum
   of 40 hr.  of training for workers at hazardous waste sites.)
•  The construction contractor estimated that the labor efficiency
   of the crews decreased by a factor of 2.5 to 3.0 for Level "C"
   protection at 60ฐF and by a factor of 3.0 to 3.5 at 90ฐF.
•  Under the Superfund law, the government provides third-party
   indemnification, with a $100,000 exclusion. The construction
   contractor included this in its bid.
•  A higher profit margin was included in the bid because of the
   work's hazardous nature. In standard construction work, profit
   margins average only 3 to 5%.
                                                                                                 SITE REMEDIATION     331

-------
  Most of these conditions will continue to be true until more con-
struction contractors are qualified to do hazardous waste work,
until the construction industry becomes more familiar with the
requirements of hazardous waste work and until firms gain the
necessary background in the training, medical  monitoring and
equipment required for this work. When these conditions are met,
a more competitive bidding environment will exist.

CONSTRUCTION  MANAGEMENT
  The on-site construction management of a remedial action can
determine whether the overall project will succeed or fail. Many
firms are becoming experienced in the RI/FS process, but few firms
have had the opportunity to implement a  remedial action.
  One of the primary concerns of the U.S. EPA during remedial
action is to ensure that the project is carried out in a manner con-
sistent with the Record of Decision. In addition, the fiscal aspects
and schedule impacts of the project are areas where the U.S. EPA
will require frequent updating. The U.S. EPA's management policy
stresses periodic site visits during the construction phase to verify
that the  activities listed below are being carried out.
  Many of the activities that will be discussed herein are not unique
to hazardous waste construction projects but are evident in any
well-managed construction project.  However,  the following
descriptions of on-site activities, the main emphasis will be on the
hazardous waste-related issues:
• Preconstruction meeting. A preconstruction meeting was held
  primarily to establish the on-site organization: procedures for
  communication, notices, change orders, payment applications
  and all other on-site administrative activities. On hazardous waste
  projects this meeting is also an appropriate time to discuss the
  health and safety plan for  the project. This discussion would
  include defining the work areas (the exclusion zone, the con-
  tamination reduction zone  and the support zone), identifying
  the decontamination procedures and detailing the personal and
  site monitoring procedures.
• Health and safety monitoring. On  the Verona site, the con-
  struction manager was responsible for the health and safety of
  all  CH2M  HILL on-site staff. Requirements from the state
  dictated that the construction manager also monitor the off-site
  air for contamination. Although the construction manager per-
  formed monitoring  both on-site and off-site, the construction
  contractor  was responsible  for  monitoring its workers on-site
  and for providing a full-time site safety officer.
• Construction progress meetings. From a management perspec-
  tive, progress meetings are important to coordinate daily activi-
  ties of all various subcontractors involved in the project work.
  On the Verona project, progress meetings were extremely help-
  ful  in coordinating  work in the exclusion  zone. For instance,
  prior  to scheduling  some work  to be accomplished in Level B
  protection, a meeting was held to discuss other on-site activities
  and to reschedule them until the Level B work was completed.
  This minimized the number of on-site workers required to dress
  in Level  B protection.
• Construction operation and testing.  This activity included rou-
  tine inspection and testing,  such as soil compaction testing and
  concrete testing. Because hazardous waste was present at  the site,
  some construction procedures had to  be modified. During demo-
  lition activities on-site, the  construction contractor had to use
  extreme caution not to violate the air quality  standards estab-
  lished by the state. The backhoe operator loading the trucks with
  debris had to  be careful not to  raise the bucket and dump the
  contaminated soil from too high above the truck; because this
  would cause the volatilization of contaminants in the  air.
• Documentation. As with all U.S. EPA-related  work, documen-
  tation is important for future cost recovery, litigation and proper
  cost justification. On this site, the documentation included the
  following:
      Daily logs of the construction activities
      Daily site safety logs
      Telephone conversation records
    - Meeting minutes
    - Correspondence
    - Soil testing reports
      Concrete testing reports
      Memoranda

• Cost documentation. Most important is the documentation of
  costs as they relate to change orders and payment applications
  from the construction contractor. The U.S. EPA requires care-
  ful documentation of any change orders,  including the following:

      Evidence of the need for a change Description of the change
      Reasonableness of the method  to implement the change
      Allowability of the costs
      An engineer's estimate of the cost of the change
      A summary of the negotiation  of  the cost
      Evidence that the change order was accomplished

• For payment applications, the U.S. EPA requires certification
  by the contractor that the work claimed has been completed.
  This completion can be certified by obtaining a schedule of values
  from the contractor that breaks the work  into components. Then,
  as the contractor submits a payment application, the construc-
  tion manager can  make  an  independent review of the work
  accomplished.
• Periodic reporting. Reporting is another  important facet of U.S.
  EPA projects. The U.S. EPA project manager must be kept wefl
  informed and up to date on the project at all times.  On the Ve-
  rona site this was accomplished  through daily reports from the
  job-site, monthly progress reports and frequent telephone con-
  versations.
• Air monitoring. This activity was specific to the hazardous waste
  portions of the work. On the Verona project, the state required
  off-site air monitoring to document  that the off-site migration
  of volatile organics did not exceed 1  ppm for more than 5 min.
  A site-specific monitoring plan was developed by CH2M HILL
  and implemented after approval from  the U.S. EPA and the
  state.  The plan called for compound-specific monitoring during
  periods of heavy site activity and hourly monitoring  around the
  perimeter of the site  daily.  During  particularly  hazardous
  activities on-site, monitoring was performed continuously at the
  nearest downwind residence to make sure the 1-ppm level was
  not exceeded.
• Schedule monitoring.  Monitoring the progress of the work is
  important on any construction project. On the Verona project,
  the construction contractor provided a 2-week schedule with
  weekly updates. This provided the construction manager with
  timely notice of any potential problems  in the work's progress.
  Without periodic schedule updates, it  is difficult to plan for
  additional effort that may be required to bring the project back
  on schedule.
• Change orders. Change orders on any construction project may
  result from unforeseen changes in conditions, owner-  or designer-
  requested changes  and contractor-suggested changes in the
  project. One of the responsibilities  of a full-time on-site con-
  struction manager is to anticipate changes and minimize their
  effect on the project. On the Verona project, the total change
  to the lump sum  bid price was less than 1.4%
• Construction closeout. Closeout of the construction  contract re-
  quired that  the construction manager prepare a final "punch
  list" of items to be completed or corrected prior to final accep-
  tance. The contractor was responsible for providing operation
  and maintenance manuals and  for training selected personnel
  in the operation of the groundwater extraction system. This is
  a critical item to specify in the contract documents and to receive
  prior to final payment.
• Construction contract administration. Administration of the con-
332    SITE REMEDIATION

-------
  tract documents was accomplished by the on-site construction
  manager, who is the person faced with the day-to-day manage-
  ment of the construction. The key to good contract adminis-
  tration is to be familiar with the contract documents from start
  to finish.
• Shop drawing submittal monitoring.  The submittal and review
  of shop  drawings is similar to any other  construction project.
  The on-site construction manager can help to facilitate the timely
  review of shop drawings by  coordinating closely with the
  designers, and the construction manager should be aware at all
  times  of the status of shop  drawing review.

LESSONS LEARNED
  Lessons learned during the design bidding and construction
phases that can be applied to other remedial design and construc-
tion projects include:

• Budgeting for remedial design and construction management
  activities requires careful consideration of many different work
  aspects.
• Firms can expect to devote considerable time to state and  local
  agencies' concerns, reviews, and general liaison.
• Firms should  consider  all of the  special requirements  of
  hazardous site conditions, special design considerations, and con-
  tractor market conditions when preparing the engineer's estimate.
During the design, the scope of work must be reviewed carefully.
Dividing the work into separate bid packages for hazardous work
and non-hazardous work may be advantageous. Many potential
bidders on the Verona project expressed interest in bidding on either
the hazardous waste portion or the off-site  construction portion,
but not both. This division then could result in more competitive
bids because more firms would  be bidding in familiar areas.
• Early in the project, contact potential bidders inform them of
  upcoming work and get an indication of how many qualified
  firms are committed to bid on the work.  It may require multi-
  ple  advertisements, phone calls and perhaps  some front-end
  preparation on the part of the construction firms.
• If possible, accept bids only from firms with a health and safety
  program and those that meet  the  requirements of 29  CFR
  ง1910.120. Inexperienced firms are a potential liability on-site
  from a health and safety standpoint. Also, the start of construc-
  tion may be delayed a month or more while the construction
  contractor fulfills the training, medical monitoring, and other
  general health and safety requirements, However, the contrac-
  tor may have to hire laborers who are not trained or under med-
  ical surveillance programs, and some lead time will be required
  for their training and physicals.
• A key factor contributing to the success of the Verona project
  was the use of a dedicated construction manager. Because of
  the multitude of daily activities and the potentially large number
  of contractors and subcontractors involved on a project, the U.S.
  EPA and state project managers  are encouraged to hire a
  qualified construction manager. Potential sources of managers
  include REM contractors, the U.S. Army Corps of Engineers
  and other engineering consulting firms.
•  A firm should be prepared to deal with construction contrac-
  tors inexperienced in hazardous waste work and be ready to bend
  in the project approach when appropriate and when conditions
  permit.
• A firm should keep thorough documentation of all aspects of
  the construction phase, particularly with respect to payment
  requests, change orders, and potential claims.
• Frequent reporting to the U.S. EPA regarding the status of the
  project and any potential problems is critical. This will help to
  ensure that the project is progressing to  the U.S. EPA's satisfac-
  tion and eliminates any surprises.

CONCLUSION
  The Verona project was successful when measured in terms of
cost and schedule control. The project was  completed and operating
within the original schedule, and the change orders on the project
were less than 1.4% of the total construction cost (as compared
to 3 to 4% on a typical construction project).
  Other firms can learn from this project experience as the require-
ments of the project design and construction differ from those of
a traditional  non-hazardous project. This remedial design  and
implementation work  can be very challenging. As more of these
projects are completed, guidance for the various work phases will
be developed. In the meantime, this paper provides some valuable
lessons learned for  your next remedial design and  construction
project—where the  action is.
                                                                                                  SITE REMEDIATION    333

-------
                                    Impervious  Liner  Installation
                                           Along Canal Bottom

                                                Eugene F.  Stecher, P.E.
                                                  Monsanto Company
                                                     St.  Louis, MO
ABSTRACT
  The placement of an 80,000 ft2, 100-mil thick  High Density
Polyethylene liner along the bottom of the Texas City Storm Water
Canal to seal residual waste products which were emitting sheens
and odors was successfully accomplished using a method not
normally used for this installation. This method allowed the liner
to be placed without complete dewatering of the canal. In this paper
the design concepts and construction methods will be described
so that this innovative method for liner placement can be used on
future projects.

INTRODUCTION
  The city of Texas City, Texas, is adjacent to Galveston Bay,
consequently its average elevation is + 5.0 ft. MSL (mean sea  level).
Because of this low elevation the city had suffered major flooding
problems during heavy rains and hurricane tide surges. To improve
this situation, the city, in cooperation with  the Corps of Engineers,
built the Texas City Flood Control System in 1964. This system
included the building of a seawall  along Galveston Bay to protect
the city from the high tides caused by hurricanes and the construc-
tion of the Storm Water Canal to improve the city's ability  to
handle drainage from heavy rains. The storm water canal would
serve as a holding pond to allow  quicker  runoff from the  City's
drainage system.
  Engineering studies dictated that the correct location for this
canal should be adjacent to the seawall. This location required the
new canal to bisect an existing hazardous waste site.  The  waste
     EXIVTIMG  WAST&  SITE
                         Figure 1
                        Plan of Site
site consisted of a series of pits that were excavated and filled with
waste chemicals from a nearby plant. After the pits were filled they
were covered with soil giving the appearance of a lush green field.
An easement was acquired from the owner of the waste site which
allowed the canal to bisect the site. After waste pits that interfered
with the canal location were excavated and transferred to adjacent
pits, the canal  was constructed.  In 1981, the city improved the
operation of the canal by adding a pump station at the end of the
canal which would pump the collected water over the seawall and
into  the bay. This station consisted of three Archimedes screw
pumps each having a capacity of 125,000 gal/min. This new pump
station kept the water level in the canal at a constant - 6.0 ft MSL
(Fig  1).
   As a result of the pump station installation and lowering of the
water elevation from + 5 to -6 ft. MSL, a hydraulic gradient was
created which caused movement  of groundwater from the waste
site through the waste pits and into the canal. As this water moved
through the pits, it carried with it portions of the pit wastes. This
leaching resulted in the appearance of floating organic material
in the canal along with a sheen atop the water which  emitted a
noxious odor.

ENGINEERING EVALUATION
   The hydraulic gradient problem which caused groundwater to
move through the waste pits and  into the canal was solved by the
installation of a slurry trench and interceptor drain along the bank
of the canal.1 However, the  sludges that were carried by the
groundwater movement still remained on  the bottom of the canal.
These sludges had to be removed in their entirety  to completely
remove sheens and odors from the canal.
   Several factors had a major effect on the ultimate solution. These
were:

•  The solution could not effect  operation of storm water canal
•  Construction work in the canal could not take place during the
   hurricane season
•  The solution must be "walk away"

CONSTRUCTION ALTERNATIVES
   The first, and most obvious solution was to sheet pile half of
the canal, dewater the area inside the sheet pile, remove the waste,
backfill with clean material and remove the piling. This procedure
would allow drainage through half of the canal while working in
the other. This solution, however, was rejected because the con-
struction period was of such length that work during the hurricane
season could not be avoided. In  addition, the construction costs
 334     SITE REMEDIATION

-------
associated with this solution were very high.
  The second solution was to dredge the material from the canal
bottom onto the waste site. This solution seemed feasible. However,
government regulations would not allow the residual water con-
tained in the dredged material to be returned to the canal without
considerable treatment. Consequently, this treatment requirement
increased the cost drastically, making this solution uneconomical.
  The third solution was to build a low water dam, above and
below the waste area in the canal. This new dam would allow water
inside the dammed area that was above the water table to be
lowered using portable pumps. The waste contained in the  silt could
then be removed by dredging this material to an area next to the
shore. The material could then be removed using a large backhoe.
The waste below the silt level that was contained in the hard bottom
and small residual amounts of sludge remaining could be sealed
with an impervious liner. If heavy rains occurred during construc-
tion, the low water dams could overflow or  be breached to allow
the water to flow through. This  solution solved the three major
concerns stated earlier but introduced a new problem. Since the
impervious liner would be placed at the bottom of the canal, its
location was well below the natural water table. Consequently, it
was impossible to keep the bottom of the canal free from water.
Therefore, the impervious liner could not be installed in accordance
with the normal installation method  of fuse welding  the liner
sections together after placement in its ultimate location.

CONSTRUCTION METHOD
  After evaluation of all possible solutions,  the installation of an
impervious liner along the canal bottom was chosen as the optimum
solution. One problem still remained; how to place the liner across
the canal which was 180-ft wide while the canal contained approxi-
mately 4 ft of water.

Silt Removal
  The contaminated  soft  sediment in the canal bottom was
removed by dredging the silty material to an area next to the shore.
A floating containment structure  was built to isolate the waste and
insure that a new area of contamination was not created. A large
backhoe located on the shore  could  reach the material in the
containment structure and place it in lined  trucks for hauling  to
the area inside  the slurry wall. This  method proved very success-
ful in moving the contaminated silt.  Approximately 15,000 yd3  of
material  were  removed from the canal in 14 days.

Liner Installation
  The 100-mil HDP liner was shipped in four rolls, each 34 ft wide
by 650 ft long.  The rolls were placed on the east side of the canal
along the seawall. The liner was unrolled parallel to the canal until
325 ft were unrolled. Since the liner was shipped in double random
lengths of 650 ft,  the liner  had to be cut.  After cutting, the
remaining 325  ft was taken  to its original location and unrolled
parallel to the section of liner that  was just placed. The first problem
to be solved was how to unroll the liner on the side of the seawall
whose slope was 3 to 1. This was handled by making  a spindle
consisting of a 2-in. diameter pipe approximately 10 ft long with
a 2 ft diameter plate welded to the end. The spindle was placed
inside the liner roll with the 2 ft diameter plate on the outside at
the low end of the roll. As the liner was unrolled, using a backhoe
with a special adapter attached to the bucket, a dozer held  the
spindle inside the roller  as it was unraveled perpendicular to  the
slope of  the seawall. (Fig 2)
  After one section of liner was rolled out and cut, the side closest
to the canal was prepared for attachment of the special rigging
needed for pulling. The  rigging consisted of eight 2-in. pipes  8 ft
long that were equally spaced along the leading edge of the liner.
The pipes were attached to the liner by threading them through
the liner. At each end of  the 2-in. pipe a 3/8-in. diameter wire rope
was attached connected  to a 6-in. diameter pipe 30 ft long. There
were four 6-in. pipes. From the end of the 6-in. pipes, a 1/2-in.
diameter wire rope was attached and brought together at a single
point approximately 40  ft in front of each pipe. From this point
a single 1/2-in. diameter drawn cable was connected to a dozer
on  the opposite side of the canal. (Fig 3.) As the leading edge of
the liner  was being rigged a second sheet was placed adjacent to
the first  and fuse welded. The rigging and the first fusion weld
took 6 hr complete. The first  section of liner was now ready to
be  pulled.
                           Figure 2
                    Section Through Seawall
                        (Looking South)
                          Figure 3
                        Liner Rigging
  Four D6 dozers were placed in position on the far side of the
canal. Attached to each dozer was the 1/2-in. diameter wire rope
from the 6-in. pipe rigging. The dozers moved in unison as the
signal was given. The liner moved down the slope toward the canal
and immediately started to wrinkle and slide at an angle. The signal
was given to stop the pull with the liner pulled only half the dis-
tance it was to travel during the first pull. A solution was needed
to solve these problems because excessive wrinkles could cause the
liner to crack. This problem was not anticipated because the ri-
gidity of the 100-mil liner samples gave the impression that wrin-
kling would not occur.
  After analyzing the problem, two things became apparent. First,
the rigging along the leading edge was not placed close enough to
the end of the liner. This caused the leading edge of the liner to
bend down slightly at random locations. This resulted in different
resistant forces along the front of the liner. Second, the liner needed
to be placed in tension during the pull.
  The amount of time required to change the rigging along the
leading edge was too  lengthy so the decision was made not to
change. Some small revisions were made using wire and duct tape
which provided enough relief so that the unequal forces could be
overcome. The tension requirement was achieved by using the
backhoe and dozer used for unrolling the liner. A cable was con-
nected to the corner of the  trailing edge of the liner and attached
                                                                                                  SITE REMEDIATION    335

-------
to each piece of equipment. As the liner was pulled, the equip-
ment connected to the trailing edge moved with the liner keeping
the liner in constant tension. This method considerably lessened
the problem of wrinkling the liner.
  Eight sections of liner, each 34 ft wide by 325 ft long were fuse
welded and pulled across the canal in 4 days. The liner was now
floating  atop of the canal. After the liner was positioned in its
correct location, it was sunk by placing 5 ft of clay atop the liner
working from the shore until the weight of the clay sank the liner.
This insured that the liner was in its proper position along the canal
bottom.  The edges of  the liner perpendicular to the canal were
secured by using tremie concrete. Small  steel anchors had been
previously placed in the liner edge to permanently secure the edge
into the concrete. The temporary dams were removed and the canal
was allowed to reach its normal depth of 15 ft. The entire operation
was completed in 2 weeks.
METHOD IMPROVEMENTS
  Two major changes in construction methods need to be incor-
porated for future installation. First, the rigging along the leading
edge of the liner must be improved. The rigging needs to be located
as near as practical to the end of the liner to help eliminate wrinkles.
The liner should be wrapped around the 2-in. pipe instead of
threading the pipe through the liner. This process will keep the
leading edge from  bending at random  locations which  causes
varying resistance forces.
  The second major improvement is the requirement for liner
tension during installation. The method  described in this report
provided the necessary tension but it was devised during the actual
installation to solve a problem that was occurring at that time.
During the planning stages for any future  installations, special
attention must be paid  to this potential problem, as  it proved to
be the most difficult obstacle to overcome during the installation.
FUTURE INSTALLATIONS
  It is apparent that this method can be applied in areas when the
water is much deeper than 4 ft. Since High Density Polyethylene
is lighter than water, the liner will float across the top of the water.
After the  liner has been pulled across  and  properly positioned,
ballast can be added to sink the liner to the bottom. The ballast
can be any material that the design engineer deems necessary for
the conditions (i.e., clay, sand, rip-rap, crushed rock, etc).
  Future applications could even  include sealing river bottoms.
Another possibility could be to reline the bottom of a storage pond
without dewatering the pond.

CONCLUSION
  When this installation method was initially proposed, bids for
assistance in design and installation were solicited  from eight firms.
Over 70^0 of the firms responded by saying the proposed method
had never been attempted before and they could foresee so many
problems  that they felt it was impossible to construct.  Conse-
quently, they declined to bid. Despite this negative response, the
installation method described in this paper proceeded.
  The successful contractor stated that he would provide construc-
tion assistance only and that the entire operation must be under
the direction of the engineer who designed this method. The design
engineer was to have full responsibility for any problems that would
occur and provide the  solution.
  The  liner has been in place for over a year. Sheens and odors
have completely  disappeared from the canal. Testing of the canal
waters has shown no trace of organics or other constituents. Tests
made to check on wildlife population  have shown a significant
increase in the fish and shrimp population.
  The  entire installation has proven  to be a complete success.

REFERENCES
1.  Stecher, Eugene F., "Remedial Action Design for Hazardous
   Waste  Landfill," presented at "Solid &  Hazardous Waste
   Management  Symposium," University of Houston, Houston,
   TX, Jan., 1985.
336    SITE REMEDIATION

-------
           Analysis of  RCRA  Closure  Options  for  Superfund  Sites
                                                John M. Cunningham
                                                   Marlene  G.  Berg
                                      U.S. Environmental Protection Agency
                                          Hazardous Site Control Division
                                                  Washington, D.C.
 ABSTRACT
  As discussed in the 1985 NCP, Superfund remedies are to be
 consistent with applicable or relevant and appropriate requirements
 of other environmental statutes. RCRA is an important environ-
 mental statute which offers several closure regulations as options
 for Superfund source control measures. The closure options, in
 turn, represent varying degrees of residual management and treat-
 ment. RCRA closure options  and associated  requirements for
 management of residuals, treatment and cleanup levels are dis-
 cussed. Additionally, the emerging concept of alternate or hybrid
 closure and its potential for increased flexibility in designing reme-
 dies that implement SARA's preference for treatment are discussed.

 INTRODUCTION
  Traditionally, there have been two closure options under RCRA:
 (1) closure by removal or decontamination and (2) closure as a land-
 fill. The premise of closure by removal (or clean closure) is that
 sufficient contamination  is removed or decontaminated so that
 there will be no significant threats to human health and the
 environment. At time of closure,  there is no containment of
 hazardous materials and no post-closure management require-
 ments.  In contrast,  landfill closure does not require removal of
 hazardous waste and therefore  allows  full  containment of
 hazardous substances. Long-term management is  required to ensure
 protection of human health and the environment.
  The Superfund program, in developing remedial actions, has
 found that combining clean closure and landfill closure require-
 ments at a number of sites has resulted in protective, cost-effective
 closures. Such a hybrid approach was used for the Crystal Chemi-
 cal Company site in Texas, as discussed in the preamble to the 1985
 NCP3.  The RCRA program has developed a third option, or
 alternate closure,4 that incorporates requirements of clean and
 landfill closure.

 CLEAN CLOSURE
  The current regulations for clean closure require the removal
 or decontamination of waste residues. If requirements are met, the
 site is no longer under regulatory control at the time of closure.
 There are no land use restrictions and no requirements for post-
 closure care.
  Conforming  changes to RCRA's interim status clean closure
 regulations have recently been promulgated.5  The  preamble to
 these conforming changes defines  what is meant  by complete
 removal/decontamination. These requirements will apply to all
closure of storage units when clean closure is elected, whether the
unit is permitted or closing under interim status.
  The amendments require that at time of closure all soils (satu-
rated and unsaturated) must achieve stringent cleanup levels.
Leachate from remaining soils must meet drinking water standards
or U.S. EPA-recommended health-based levels. Soil contaminant
levels must be below those levels established by the U.S. EPA as
acceptable for inhalation, ingestion and dermal contact. Natural
attenuation is not presumed.  Clean closure is best suited for sites
where hazardous wastes are easily located and removed and  are
of limited quantity.
LANDFILL CLOSURE
  Landfill closure allows full containment of hazardous material.
To minimize the potential for release, requirements2 include im-
permeable covers and long-term management of the site and cover,
maintenance of leachate collection and removal systems, ground-
water monitoring and, if necessary, corrective action. Covers must
provide long-term minimization of migration of liquids through
the closed landfill and have a permeability less than or equal to
the permeability of any bottom liner system or natural subsoils
present. Land use is restricted during the post-closure period. Land-
fill closure may be selected where hazardous wastes are scattered
over  a wide area or there are large volumes of wastes.

ALTERNATE CLOSURE
  In  addition to the two closure options described above, the U.S.
EPA has proposed to add a third closure option.4 It is known as
hybrid or alternate closure. The basic premise of the hybrid or
alternate closure is discussed in the preamble to the proposed hybrid
closure rule:

      "The Agency now has several years of experience in
   reviewing and approving closure plans under RCRA.
   Moreover, the Agency has gained considerable experience
   in effecting remedial actions under the Comprehensive En-
   vironmental Response  Compensation and Liability Act
   (CERCLA) that are in many ways analogous to RCRA
   closures. Based on this experience, EPA believes that in
   many circumstances a "hybrid" approach that combines
   the strategies of closure by removal and closure as a dis-
   posal unit may be equally or more effective than either the
   pure "disposal" or "removal" closure option. Rather than
   designing all caps to minimize infiltration and allowing the
   waste to remain in place, this "hybrid" approach would
   consist of the removal of the majority of contaminated
   materials and would allow covers and post-closure monitor-
   ing to be designed based on the exposure pathway of con-
   cern (Ibid at 8713)" (Emphasis added).
                                                                                            SITE REMEDIATION    337

-------
  Under the proposed alternate closure, some hazardous waste con-
stituents could remain at a site and a closure plan tailored to the
specific pathway(s) of concern would be permitted.  A series of
factors is provided in the proposed rule that a Regional Adminis-
trator would be required to assess before approving an alternate
closure plan. The alternate closure option may be selected where
potential pathways of exposure are limited  and where remaining
contamination is  low in mobility or toxicity.
  The alternate closure approach was proposed Mar. 19, 1987.*
The U.S. EPA expects to promulgate a final  regulation in May,
1988.

EXAMPLES OF ALTERNATE CLOSURE
  The proposed rule for alternate closure4  includes a number of
scenarios based on routes of exposure.  The Superfund program
has focused on two scenarios or types of alternate closure. In one
scenario, which has been designated "alternate-clean"  closure, the
majority of waste material is removed.  Residual contamination,
however, does not meet clean closure requirements. In this case,
based on potential threats, no containment is required and post-
closure care is minimal. A second scenario, designated "alternate-
landfill" closure, involves the removal  of wastes  such that con-
tainment is required. However, based on the  potential pathways
of exposure, a permeable cover suffices and post-closure care is
less rigorous than that for landfill closure.
  In the case of "alternate-clean" closure (Figure 2), residuals pose
no  direct contact threat nor  impact on groundwater. Residual
leachate contaminant levels exceed health-based levels. A cover is
not needed to  ensure that the site is protective of human health
and the environment. However, the proposed rule requires  fate
and transport modeling and model verification to ensure that the
groundwater aquifer is usable. While no containment or long-term
management is required, modeling must be  conducted and a deed
notice is recommended.
                   MATERIAL MANAGED ON-SITE

-------

Description
Cleanup Criteria
Analysis needed lo
demonstrate closure
requirements have been
met
Engineering required
long-term management
Situations where most
useabke
Benefits/Disadvantages
to Owner/Operators and
Responsible Parties
t-5
CLEAN CLOSURE1
• All wastes removed
• No containment or management
controls
• Leachate at health-based levels
• No significant direct contact
threat
• Assume no natural attenuation
• Fate and transport modeling is not
used
• Confirmation monitoring of soils
and ground water
None
• No long-term management
needed
• Sharp delineation of wastes
that are easy to identity,
remove and treat
• No widespread subsurface migra-
tion of hazardous constituents
Benelits:
• No post-closure program
* Unrestricted land use at closure
Disadvantages:
• Requirement for substantial
removal
• Stringent dean up levels
ALTERNATE
"ALTERNATE
CLEAN CLOSURE"4
Majority of wastes removed
No containment
Model verification
No long-term management
No significant direct contact threat
Leachate exceeds health-based levels
No significant threat to ground water
Exposure based risk assessment
Soil analysis to demonstrate no significant
direct contact threat
• Fate and transport modeling.verification of
model prediction lo ensure ground water
aquifer is usable
None
* Model verification for ground water
• Deed notice (recommended by Superfund
program)
CLOSURE 2
"ALTERNATE
LANDFILL CLOSURE1*4
• Partial removal of wastes
• Limited containment (i.e. soil cap)
• Minimal long-tern) management controls
* Leachate at health-based levels, no
significant threat to ground water
• Cap required to address dired contact threat
• Exposure based risk assessment
• Minimal ground water monitoring
Soil cap (may be permeable)
• Limited long-term management: maintenance
of site, cap
• Land use restrictions
• Minimal ground water monitoring
• Pathways of potential exposure of contaminants is limited
• Remaining contamination has low mobility
• Remaining contamination has low toxidty
Benefits:
* Partial removal/decontamination of wastes
• Do not have to meet soil/water dean-closure requirement levels
• Cover, post-closure program are site specific, tailored to potential pathways of exposure
Disadvantages:
• Some land restrictions; limited post-closure program: maintenance of site and cap; model
verification
LANDFILL CLOSURE3
• No requirement for waste removal
• Containment of hazardous substances
• Long-term management controls
• Significant management and maintenance
controls to ensure protection/no releases
• Monitor ground water to determine If landfill
Is releasing hazardous substances
• Monitor air, surface water, soils to evaluate
integrity of cover
Stabilize waste to support final cover, Install
impermeable cap, teachate collection system
• Long-term management of cap and site
• Pump and treat leachate, monitor ground
water, and if necessary, conduct corrective
action
• Land use restrictions, survey plat to be filed
with county, deed notice required
* High concentration of contaminated soil/
wastes
• Large volumes
• Treatment/removal is not pradical/cost-
effective
Benefits:
• No treatment/removal of waste
Disadvantages:
• Need for impermeable cap
• Long-term management
* Land use rest net ions, survey plat, deed
notice
     1  40 CFR Part 265. Interim Status Standards for Owners and Operators of Hazardous Waste Treatment, Storage, and Disposal Facilities; Final Rule. March 19,1987.

     2  40 CFR Parts 264, 265 and 270. Proposed Amendments for Landfill, Surface Impoundment and Waste Pile Closures; Proposed Amendments to Rule. March 19,1987.

     3  40 CFR 264.310, 265.310. New landfill units are to be retrofitted with double liners to meet minimum technology requirements, 40 CFR 264.301; 265.301.

     4  Based on the proposed alternate closure rule, two examples of types of alternate closure are provided. One, which is similar to dean closure, has been given the name 'Alternate Clean Closure.'
       The second one. designated 'Alternate Landfill Closure,' is more similar to landfill closure.

                                                               Figure 2
                                                    Hazardous Waste Closure Options
might be determined not to be relevant and appropriate since they
are designed  to  prevent  groundwater contamination problems
resulting from leaching of hazardous constituents from the wastes.
The fate and transport modeling show that such contamination
is not a problem sufficiently similar  to that encountered at the
CERCLA site.
  By contrast, the monitoring requirements might be determined
to be relevant and appropriate since  they are designed to detect
groundwater contamination and at this site such monitoring may
be required to confirm the accuracy of the fate and transport
modeling. A similar analysis would be conducted for other clean
closure and landfill closure requirements and other potential routes
of exposure. Such an analysis may, for example, lead to the con-
clusion that a cover other than a full RCRA landfill cover is neces-
sary (i.e.,  to  protect against potential direct contact threats).

MANAGEMENT OF RESIDUALS
  In determining which closure options are to be used,  an addi-
tional factor needs to be considered. This factor is the applicability
of the Land Ban requirements. Land Ban requirements are applica-
ble when restricted wastes are "placed" after a relevant  effective
date. Placement is analyzed by reference to an area of contamina-
tion in a way similar to  the analysis  of disposal.
  If the Land Ban requirements must be  met,  RCRA  hazardous
wastes  are treated to BDAT levels. In turn, BDAT residuals are
considered RCRA hazardous wastes and as such are regulated
under the RCRA Subtitle C program. The range of residual BDAT
levels will influence the type of closure to be used under Subtitle
C. For example, a few BDAT levels will be as low as clean closure
levels. These materials could be removed from regulatory control
by utilizing a clean closure.
  Sites containing treated material posing no groundwater threat,
but posing a direct contact threat, could be closed using "alternate-
landfill" closure. These materials also could be delisted. Delisted
materials must pose no groundwater threat 500 ft from the point
where the waste enters the groundwater. However, delisted wastes
may pose a direct contact threat. Under the current RCRA delisting
regulations  (40 CFR  Sec. 260.22), delisted materials require no
further management. However, the U.S. the U.S. EPA is con-
sidering requiring further management under Subtitle D. BDAT-
treated residuals  above delisting or "alternate landfill" closure
levels would have to  undergo landfill closure.
  An array of closure options and associated clean up levels are
provided in Figure 3 for BDAT-treated residuals as well as for
materials not affected by the Land Ban.
 CONCLUSION
   RCRA closure options offer several levels of residual manage-
 ment for CERCLA remedial actions. Associated with each closure

                                SITE REMEDIATION     339

-------
                             Treat RCRA Hazardous
                             Waste to BOAT Levels
                     Disposal On-Site*
    1. Clean Closure:
        Soils, leachate are at health-based levels;
        no further regulatory control.

    2. Alternative-Clean Closure:
        No direct contact or ground water threat.

    3. Alternate-Landfill Closure:
        Direct contact threat, no ground water threat.

    4. Deli sting":
        No ground water threat 500 ft from contaminant
        point of entry into ground water. Possible direct
        contact threat. Currently, no further management.

    5. Landfill Closure:
        Potential direct contact, ground water threat.
    Under RCRA Subtitle C regulatory control.
    Off-Site Disposal:
    1) delisted material - no further management now.
    (U.S.  EPA is considering Subtitle D control), and
    2) materials above delisted levels must go to C facility.

    U.S. EPA is considering further management under RCRA
    Subtitle D.
                         Figure 3
           Choices for Hazardous Waste Management
option are requirements for treatment and cleanup levels. The appli-
cability of the Land Ban requirements wiil influence what cleanup
levels are to be achieved and thus which closure options are avail-
able for use. The availability of alternate closure options at Super-
fund sites provides greater flexibility in allowing the U.S. EPA to
more effectively implement the goals of SARA including the use
of treatment.
REFERENCES
I.  40 CFR Sec. 264.197,258; 40 CFR Sec. 265.197,228,258.
2.  40 CFR Sec. 264.310; 40 CFR Sec. 265.310.
3.  "National Oil and Hazardous Substances Pollution Contingency
   Plan. Final Rule." U.S. EPA, Nov. 20, 1985.
4.  40 CFR Pans 264, 265 and 270. "Proposed Amendments for
   Landfill, Surface Impoundment and Waste Pile Closures; Pro-
   posed Amendments to Rule." U.S. EPA, Mar. 19,  1987.
5.  40 CFR Pan 265. "Interim Status Standards for Owners and
   Operators of Hazardous Waste Treatment, Storage and Dis-
   posal Facilities;" U.S. EPA,  Final Rule. Mar. 19, 1987.
340    SITE REMEDIATION

-------
            An Innovative  Approach  for Remediation of Metal-
       Contaminated and  Environmentally Sensitive  Marsh  Areas

                                          Hsin H.  Yeh,  Ph.D., P.E.
                                        Dev R. Sachdev, Ph.D., P.E.
                                        Ebasco Services Incorporated
                                            Lyndhurst, New Jersey
                                              Joel  A.  Singerman
                                   U.S. Environmental  Protection Agency
                                Emergency and Remedial Response Division
                                            New York,  New York
ABSTRACT
  Feasible remedial alternatives were developed for the cleanup
of the metal, contaminated (Cd, Ni and Co) sediments in the East
Foundry Cove Marsh/Constitution Marsh portion of the Marathon
Battery Company Superfund site located in Cold Spring, New
York. The remedial alternative selected by the U.S. EPA for East
Foundry Cove Marsh was "Hydraulic Dredging/Thickening/Fixa-
tion/Off-site Disposal/Revegetation,"  while  for Constitution
Marsh, "No Action" was selected. "No Action" for Constitution
Marsh was preferred due to the excessive detrimental environmental
effects that would have occurred during remediation and the
anticipated gradual improvement through natural processes after
cleanup of East Foundry Cove Marsh, which is the primary source
of contamination.
  Extensive bench-scale treatability tests were performed to
confirm the applicability of the remedial technologies including
hydraulic  dredging, gravity thickening  and chemical fixation.
Special dredging schemes were developed to minimize resuspen-
sion and redistribution of contaminated sediments as well as to
maintain the tidal exchange between the site and the Hudson River
to support the biota at the site during remediation.

INTRODUCTION
  The Marathon Battery Company Superfund site, located in the
Village of Cold Spring, Putnam County, New York, approximately
40 miles north of New York City, includes  a former battery
manufacturing facility and the surrounding plant grounds, the
Hudson River in the vicinity of Cold Spring and a series of river
backwater areas known as Foundry Cove and Constitution Marsh.
  Foundry Cove, a shallow bay and cattail marsh on the east bank
of the Hudson River across from West Point, is composed of east
and west components. East Foundry Cove is partially isolated from
West Foundry Cove and the Hudson River by a railroad bed to
the west. The 48 acres East Foundry Cove consists of approximately
14 acres of cattail marsh (East Foundry Cove Marsh) and 34 acres
of tidal flat and cove. The exchange of water between East Foundry
Cove and West Foundry Cove during flood and ebb tides occurs
through a 30 feet passage under a railroad trestle and a channel
and dike system which connects Foundry Cove to Constitution
Marsh (CM), a 270 acres wildlife sanctuary owned by the Taconic
State Park Commission and operated by the National Audubon
Society for the purpose of nature observation.
  Periodically between 1952 and 1979, East Foundry Cove Marsh
received waste effluent from the former battery plant via the
Kemble Avenue storm sewer  (see Figure 1). Both cadmium (Cd)
                        Figure 1
                  Marathon Battery Site
        East Foundry Cove Marsh and Constitution Marsh

and nickel (Ni) were used in large quantities at the battery manufac-
turing facility, and at one time cobalt (Co) was used as an additive.
It has been estimated that 56.2 short tons of suspended and l.U
short tons of soluble cadmium were discharged into East Foundry
Cove Marsh (ECM) during the course of the plant's operation.
Consequently, East Foundry Cove  Marsh sediments are con-
taminated with heavy metals of concentrations up to 17.1% Cd,
15.6% Ni and 0.6%  Co.
  In order to investigate the feasibility of cleaning up contamina-
tion at the Marathon Battery Company site, the New York State
Department of Environmental Conservation (NYSDEC), utilizing
funds provided through a cooperative agreement with the U.S. EPA
contracted with Acres International Corporation in 1984 to perform
                                                                                       SITE REMEDIATION    341

-------
a remedial investigation/feasibility study (RI/FS). In August 1985,
draft RI and draft FS reports1'2 which called for further infor-
mation to be gathered before development of a final RI/FS report
and the selection of appropriate remedial measures were completed.
In April 1986, the U.S. EPA authorized Ebasco Services Incor-
porated (Ebasco) to conduct a supplemental RI/FS at the site. RIJ
and FS Reports4 for the East Foundry  Cove Marsh/Constitution
Marsh portion  of the site were completed  in August 1986.
  This paper presents the results of the  1986 FS report4, which
evaluated innovative and conventional technologies for the remedi-
ation of the metal-contaminated and  environmentally  sensitive
marsh areas in question. The  results from the  RI reports'-' are
referenced only as necessary to support the evaluation bases of var-
ious alternatives being discussed in this paper.

EXTENT OF CONTAMINATION
  The results from the RI1'3 indicate that the concentration pro-
files of Cd, Ni and Co in the sediments of East Foundry Cove and
Constitution Marsh show similar distribution patterns. These pat-
terns reflect the origination of point source discharges of these me-
tals  from the plant into East Foundry  Cove Marsh where the
concentrations are the highest, their subsequent transport into East
Foundry Cove and adjacent Constitution Marsh, and their gradu-
ally  decreasing concentrations with distance  from the discharge
point. Generally, Cd has slightly higher concentrations than Ni,
and Go's concentrations are two to three orders-of-magnitude lower
than those of Cd and Ni. In addition,  the contamination is con-
centrated in the surficial sediments at depths no greater  than 14
in. in East Foundry Cove Marsh and of  no greater than 4 in. in
the other areas (see Table  1).
                           Table 1
      Extent of Cadmium Contamination In East Foundry Cove
               And Constitution Marsh Sediments

Cd Concentritlon (M/ka)
Location RiQฃf Avenue
Cist Foundry 100 to 170,000 26.000
Cove Hirsh
(ฃCH)
lift Foundry 10 to 60.000 1.100
Cove Tldll
Flit
(ECTF)
Constitution 4 to 940 1(0
MlrsH
(CM)
Aril
Depth to Enclosed by
dckground 100 "g/lj Totil
Cd Lfvel liopleth Arei
an iim uui uu ttuj uu
36 14 $.1 12.5 5.7 14


10 4 I.S 21.0 13. ( 34



10 4 40.5 100 109.3 270


  Kott:  Since, uong the three Mtlll. Cd It thl Kit toxic Ind III concentritloni ปen
       gmerilly tlป Mghfit. the public httltli ivtluttton utiltied tMl Mtll.
REMEDIAL RESPONSE OBJECTIVES
  Based on the results of the RI, a wetlands assessment and a public
health and  environmental evaluation, the following  remedial
response objectives were established:

• Prevention  of all  biota  from contacting East Foundry  Cove
  Marsh and  Constitution Marsh contaminated sediments that
  would threaten them
• Prevention of resuspension and redistribution of the contami-
  nated sediments that would threaten the area flora and fauna
• Minimization of the disturbance to Constitution Marsh, since
  this wetland is a delicate ecological  habitat

  In the absence of  standards or criteria for contaminant  levels
of Cd, Ni and Co in sediments to evaluate remedial alternatives
for East Foundry Cove Marsh  and  Constitution  Marsh, it was
necessary to establish an acceptable  cadmium contaminant level
for the site.
  Although Cd, Co  and Ni are found  in elevated concentrations
in the flora and fauna of East Foundry Cove and Constitution
Marsh, the marsh ecosystem is generally considered to be healthy.
In addition, no major ecosystem-wide impact which could have
resulted from a stressed environment has been found. Based upon
a probabilistic human health impact  assessment prepared by
Ebasco, however, the following conclusions were drawn:

• Site biota are contaminated with Cd to the extent that indiscrete
consumption of certain species presents a human health risk, with
the area of greatest health concern being East Foundry Cove
• Constitution Marsh presents only a minor health impact
• Sediments containing less than or equal to 900 mg/kg Cd pose
  no  health-related concerns

  Since, among the three metals, Cd is the most toxic and its con-
centrations were generally the highest, the public  health evalua-
tion utilized this  metal.
  Based upon an analysis of available information and research
data as well as discussions with state and federal fish and wildlife
experts, a site-specific  sediment Cd  remediation level of 100 mg/kg
was established. No similar levels were established  for Ni and Co
since  it was believed that  any remedial  action undertaken based
on Cd action  level would simultaneously mitigate Ni  and Co con-
taminant levels.

BENCH-SCALE TREATABILITY TESTS
  In order to evaluate the applicability of potential remedial tech-
nologies, laboratory bench-scale tests were conducted. These tests,
which included elutriate, settling, filterability, acid leaching, base
recovery and chemical fixation tests, are discussed below.

Elutriate Test
  The objectives of the elutriate  test, which was conducted to
simulate  the  mixing of sediment  and site water caused by the
hydraulic dredging operation, were:

• To  determine whether Cd, Co and Ni would be leached out or
  remain associated with resuspended solids
• To  determine  whether  the ambient  water quality would be
  degraded
• To  determine whether the dredging water would require treat-
  ment prior to its discharged back to the  site

                           Tabk 2
         Comparison  of Water Quality Between Elutriate
                Test Superaatant and Site Water
                                                                            flrซปUr
  tut Foundry Caw

  D1IUlปM U, ig/l
  Dtiulmd Co, lt/1
  D1lio)ปtd "1. 19/1

  Totll Cd. ซg/l
  lottl Co, ซg/l
  Tout HI. ซ9/l

  WS. K/1
  TurbldHy. NTU
  IOC. •)/)
  00, *g/l

  pM, unit
  AUillnlty 111 CiCO])ig/l

  Constitution MirปH
0.032
0.011
0.045

J.OSO
0.102
l.SSO

1428
 952.9
 •2.3
  3.0

  7.66
 89.0
                                                                                                           VM
                                                                                                         Stdtoxit
O.OM
0.011
0.2M

3.600
0.1ป
2.2*

2S96
19(0
 156.4
  1.*

  7.H
  98.7
O.OM
0.011
0.057

C.OOO
0,20!
4.380

2833
2453
 19$.)
  0.7

  7.65
 104.5
                                                         Silt
                                                         UlUr
 .0)3
 .811
 .OB

 .13f
 .011
 .OR
11.41
 7.85
 7.M
73.13
DIltolvM
IMl olvtd
Oil olvta
Tot 1 Cd,
Tot 1 Co.
Tot 1 NI.
TSS. ซJ/1
Turtldltv
IOC, m/1
DO. •)/!
pN, unit
Cd, cg/l
Co. ซo/l
Ml. •"/!
•1/1
•B/l
•g/i

, NTU



* Hi) Inlt, (11 CiCOjlซg/l
0.0(4
0.037
0.047
O.S53
O.OM
0.806
360?
24(3
147.2
J.38
t.S7
15.3
0.035
0.02S
O.OM
1.265
0.186
1.792
8330
577$
299
2.35
t.59
58.36
0.038
0.030
0.068
2.0(8
0.171
2.437
1)50)
8900
4)9.8
1.5(7
6.55
76.76
.0053
.011
.0215
.014
.014
.MIS
28.95
9.64
M7
7.S
7.)
53.0
                                                                      Notll:  In flutrlltt ttlti. tdt ltdlMnti w. •tisurtd .1 ptrctnt by volw.
342     SITE REMEDIATION

-------
  Three sediment samples each from East Foundry Cove and Con-
stitution  Marsh were used in the elutriate tests. Tests on each
sediment sample were performed at sediment-to-water ratios of
10%, 20% and  30%, sediment by volume.
  Table 2 shows the comparison of water quality between elutriate
test supernatant and site water. Table 3 presents the distribution
of Cdy Co and Ni after hydraulic dredging. These tables indicate
the following:

• Most of the Cd, Co and Ni released from the sediment during
  hydraulic dredging would be in with  resuspended solids
• The levels of  dissolved metals in  the  supernatant or dredging
  water are slightly higher but not  substantially different from
  those in the site water
                            Table 3
        Distribution of Sediment Metals After Elutriate Test
  Parameter

  East Foundry Cove Sediments

  Cadmium


  Cobalt
  Constitution Harsh Sediments

  Ctdltin
  Cobalt
  Nickel
                                Percent Metal
                                Leached
                                    (I)
                              Percent Metal
                   Percent Metal  Remained in
                   Resuspended   Sediment
                      ispended
10
20
30

10
20
X

10
20
30
10
20
30

10
20
30

10
20
30
0.031
0.019
0.012

0.0
0.034
0.0

0.015
0.006
0.003
0.12
0.11
0.06

0.59
0.44
0.30

0.24
0.28
0.34
3.69
3.00
2.21

3.60
0.92
2.13

4.23
1.86
2.33
4.79
4.88
4.84

5.21
5.50
5.18

6.82
6.46
5.48
96.28
96.98
97.78

96.4
98.74
97.87

95.76
98.13
97.67
95.10
95.02
95.10

94.20
94.07
94.53

92.93
93.26
94.18
  The supernatant (or dredging water) from the hydraulic dredging
 operation may degrade ambient water quality. This water quality
 impact can be minimized by removing the suspended solids from
 the dredging water.

 Settling Test (or Zone Settling Rate Test)
  The purpose of the settling test was to determine the settleability
 of the dredged material to assist in the adequate sizing of sludge
 removal basins to be used in hydraulic dredging operations for gross
 removal of solids. Due to the limited  space available at the site,
 the size of the holding basin could be a limiting factor on the
 dredging methods and rates.
  Five sediment samples each from East Foundry Cove and Con-
 stitution Marsh were used in the zone settling rate test. A hydraulic
 dredging of 20% sediment by volume was assumed in the test. The
 results of the test indicate the following:

 • The settling rates of the dredged material are high. Most solids
  settle in 30 to 60 min
 • The maximum area required for the holding basin (or thickener)
  would be 8,300 ft2 for a hydraulic dredging flowrate of 1,000
  gal/min.
 • The holding basin (or thickener) can thicken the solids from an
  initial concentration of 5 to 6% by weight to 14 to 16% by weight

 Filterability Test
  The filterability test was carried out to determine the dewatera-
 bility of dredged sediments by mechanical means (e.g., vacuum
 filters). The data provided  information to calculate the specific
 resistance of the dredged sediments which can, in turn,  be used
 to size the vacuum filters and other dewatering equipment. This
test was a continuation of the settling test discussed above.
  The test results showed that vacuum filtration can dramatically
reduce the water content of previously settled sludge (or dredged
material) from an average of 18-20%  to 37-39% by  weight. The
average specific resistance was calculated to be 2.93 x 10" secVlb
and 1.94 x  10" secVlb for East Foundry Cove and Constitution
Marsh dredged material, respectively. These values are considered
favorable for filtration without the use of any conditioning agents.

Acid Leaching Test
  The primary purpose of the acid leaching test was to obtain per-
formance data on  the extraction of metals from sediments for
resource recovery. In addition, the test results would also indicate
whether the contaminant metals  in the sediment would mobilize
under acidic conditions which could result from anaerobic condi-
tions developed by certain remedial alternatives  involving  con-
tainment.
  The acid leaching test consisted of exposing the sediments to acid
solutions of pH 2.0,3.0 and 4.0. Three sediment samples each from
East Foundry Cove and Constitution Marsh were used for the test,
and HC1 was used as the extraction solution.
  Table 4 shows the distribution of Cd, Ni and Co in the sediment
and solution for different pH levels. As expected, more metals were
leached out from the sediment at lower pH values. However, even
at pH of 2.2 to 2.5, the maximum average percentages of Cd, Ni
and Co leached out from the sediment were only 27.0%, 12.0%
and 6.0%, respectively which are not good enough for resource
recovery.  The majority of  these metals  still remained in  the
sediment.

                            Table 4
     Distribution of Sediment Metals AFter Acid Leaching  Test
                                                                         Parameter

                                                                         East Foundry Cove Sediments
                                                                         Cobalt
                                                 Nickel
                                                                                             Leปel
                        2.2-2.5
                        3.1-3.3
                        3.8-4.0

                        2.2-2.5
                        3.1-3.3
                        3.8-4.0

                        2.2-2.5
                        3.1-3.3
                        3.8-4.0
                                                 Constitution Harsh Sediments

                                                 Cadmium
                                                 Cobalt
                                                 Nickel
                                                           2.3-2.5
                                                           3.2-3.4
                                                           3.7-3.8

                                                           2.3-2.5
                                                           3.2-3.4
                                                           3.7-3.8

                                                           2.3-2.5
                                                           3.2-3.4
                                                           3.7-3.8
                                                                                                          Percent Metal
                                                                                                          Leached
27.3
 5.58
 2.46

 5.18
 5.94
 6.25

11.81
 0.98
 4.34
                                                              6.75
                                                              3.83
                                                              0.95

                                                              2.94
                                                              2.14
                                                              2.30

                                                              6.08
                                                              1.97
                                                              2.50
                                                  Percent Metal
                                                  Remained In
                                                  Sediment (ป)
72.7
94.42
97.54

94.82
94.06
93.75

88.19
99.02
95.66
                                                                   93.25
                                                                   96.17
                                                                   99.05

                                                                   97.06
                                                                   97.86
                                                                   97.70

                                                                   93.92
                                                                   98.03
                                                                   97.5
                                              Fixation Test
                                                The purpose of conducting the fixation test was to confirm
                                              whether the Cd, Ni and Co in the sediments can be chemically stabi-
                                              lized or physically bound to the sediments such that total teachable
                                              metals are reduced to levels below those listed in the RCRA EP
                                              toxicity test. If successful, the contaminated sediments would be
                                              considered non-hazardous and could be suitable for disposal in
                                              a non-hazardous waste landfill.
                                                Two types of chemical mixtures were added to the sediments
                                              to achieve chemical fixation of  the contaminant  metals.  One
                                              mixture consisted of pozzolan (from fly ash) and lime (formulated
                                              by Associated Chemical and Environmental Services , and the other
                                              consisted of sodium silicate solution and Portland cement (formu-
                                              lated by Chemfix Technologies, Inc.). The results indicate that both
                                              chemical mixtures can chemically fixate the contaminant metals.
                                              The final product volume resulting from addition of fly ash and
                                                                                                     SITE REMEDIATION     343

-------
lime, however, was more than twice the sediment volume, while
the final product volume resulting from addition of sodium silicate
solution and Portland cement was only 1.13 times the sediment
volume.

IDENTIFICATION AND SCREENING OF REMEDIAL
TECHNOLOGIES AND  ALTERNATIVES
  General response actions were identified that addressed the site
problems, achieved the response objectives and met cleanup goals.
These included removal, isolation or containment, and treatment.
Possible technologies associated with each response action were
then identified and investigated. Table 5 summarizes the feasible
response  actions and associated remedial technology categories.
Under these categories, more than 70 technologies were identified.
                            Tables
              Feasible General Response Actions And
                 Associated Remedial Technologies
    General Rttponst
    	Actions
    Contitwnt
    Control

    Hydraulic Control

    Complete Rtmval


    Partial ซซoval


    On-SItt TrtatMnt

    Off-SItt TrntMKt

    In Situ TrtltMnt

    On-SItt Disposal

    Off-sin Disposal


    Transportation

    Sltt itstontlon
                                TfCfi|f[fl]MV CjattgQr I tl

                                HonUorlng. Reitrlettd Acettt.
                                Public Anrtnttt

                                (Upplng, Stdlwnt Dliptrtlon Control
Tloal Control.  Runoff Control

Dredging (MChanlcal. hydraulic, pneutia-
ttc). Etcivttlon

Dredging (Btclunlcal. hydraulic, pntuaa-
tlc). Eicivatton

Thermal, Chtซ1cal, and Physical Trtatatnt

Thtnul. Chnlcal, and Physical Trtatatnt

Chnlcal and Physical Trtatwnt

landfill (hazardous or non-hazardous)

Landfill (hazardous or non-hazardous),
octan disposal

Truck, Train, Bargt, Plptllnt

Harsh Revtgttatlon
  The identified remedial technologies were screened to eliminate
those which were unproven technologies, those which were found
to be infeasible or inapplicable based upon the results of bench-
scale treatability tests or those which would not achieve remedial
objectives.
  The technologies which passed the screening were used alone or
in combination to developed a list of potential remedial alterna-
tives. The following remedial alternatives were identified for East
Foundry Cove Marsh (ECM) and Constitution Marsh (CM):

• Alternatives ECM-1 and CM-I: No Action
• Alternatives ECM-2 and CM-2: Hydraulic  Dredging/Thickening/
                               Fixation/Off-site Disposal
• Alternatives ECM-3 and CM-3: Hydraulic  Dredging/Thickening/
                               Fixation/On-site Disposal
• Alternatives ECM-4 and CM-4: Hydraulic    Dredging/Thicken-
                               ing/Dewatering/Off-site Disposal
• Alternatives ECM-5 and CM-5: Containment

Base Recovery Test
  The purpose of the base recovery test was to recover  the con-
taminant metals leached out by acid extraction through precipita-
tion at higher pH  levels. This test was a continuation of the acid
leaching test. NaOH  hydroxide (NaOH) was used to increase  the
pH to 7,9 and 11. The results indicate that greater than 98% and
99%, of the metals in the acid leaching solution can  be recovered
by adjusting the pH  to the levels of 9 and 11, respectively.
  Alternatives 2 through 4 were derived from the basic remedial alterna-
tive consisting of removal, treatment and disposal technologies. These  five
potential remedial alternatives were identified for East Foundry Cove Marsh
and Constitution Marsh.
                                         The potential alternatives were evaluated on the basis of the established
                                       criteria  including  technical feasibility, environmental  impact and cost.
                                       Through this screening process. Alternatives CM-3 and CM-5 for Consti-
                                       tution Marsh were eliminated from further  consideration on the basis of
                                       their severe negative environmental impacts on Constitution Marsh. On-
                                       site disposal (Alternative CM-3) and containment (Alternative CM-5) would
                                       result in a permanent loss of, respectively, 17 acres and 100 acres of marsh
                                       area in Constitution Marsh. The disturbance and modification of the marsh
                                       would promote invasion of the remainder of Constitution  Marsh by less
                                       desirable plant species and would destroy the current integrity of the Marsh-
                                       woodland ecosystem. No potential remedial alternative was eliminated based
                                       on the initial cost.  For East Foundry Cove Marsh, all the five alternatives
                                       were retained for  further detailed evaluation.

                                       DETAILED EVALUATION OF REMEDIAL ALTERNATIVES
                                         A preliminary conceptual engineering  design and description of each
                                       remedial alternative including major facilities/equipment and construction
                                       components were performed to provide the  basis for detailed evaluation.
                                       This evaluation, carried out according to NCR Section 300.68(h), discussed
                                       the cost-effectiveness of an alternative in terms of technical, environmental,
                                       public health and institutional concerns. Table 6 presents the criteria used
                                       for detailed evaluation of the remedial alternatives. In the  following, the
                                       conceptual design  of some important remedial technologies incorporated
                                       into the alternatives is described prior to the presentation of evaluation
                                       results.
                    Table 6
           Detailed Evaluation Criteria

o    TECHNICAL FEASIBILITY

         Performance
     -   Reliability
     -   Impleraentablllty
     -   Safety
     -   Level of Remediation  Achievable

0    ENVIRONMENTAL CONSIDERATIONS

     -   Beneficial effects
         Adverse effects

o    INSTITUTIONAL CONSIDERATIONS

         Confortnance  to the ARAR
         Permitting requirements
         Legal constraints, 1f any
         Cultural Resources

o    PUBLIC HEALTH CONSIDERATIONS

         Minimization of exposure
         Minimization of chemical releases
         Releases that will not be minimized
         Exposures during remedial action
         Exposures after remedial action
                                                   COST
                                                        Capital  cost
                                                        Operation and  maintenance costs
                                                        Present  worth  cost
                                      Description of Conceptual Design

                                      No Action
                                         No action  is not a  category of technologies but a group of
                                      activities which can be used to address the contamination problem
                                      when no remedial measures will be implemented. Adoption of a
                                      no-action approach does not preclude implementation of techni-
                                      cally simple measures designed to monitor potential risks over time
344    SITE REMEDIATION

-------
or to limit public exposure to site contaminants. The measures
considered were the installation of a security fence and warning
signs and long-term monitoring. In Alternative ECM-1, diversion
of storm sewers also was considered to minimize inflow to East
Foundry Cove Marsh to minimize resuspension of highly contami-
nated sediments in East Foundry Cove Marsh.

Hydraulic Dredging
  Portable horizontal auger-cutter dredges were considered  for
removing the contaminated sediments in view of their capability
in shallow-water  application,  accessibility  to the  site  and low
sediment resuspension. For dredging operations, two important
environmental concerns must be considered:  minimizing resuspen-
sion and redistribution of contaminated sediments and maintaining
the tidal exchange between the site and the Hudson River in order
to support the biota at the site during remediation.
  Special dredging schemes were developed. For the marsh areas,
the dredging scheme involves staged ponding (diking and flooding)
of contaminated sections and dredging within the ponds by using
a portable  hydraulic dredge. Earth dikes would be used to build
the ponds, and water would be pumped from the Hudson River
to flood the ponds to maintain approximately 3 ft. depth for dredge
mobility. Since East Foundry Cove Marsh area is relatively small
and is not located in the main course of tidal flows, only one pond
was considered for this  area.
  However, four ponds were considered for Constitution Marsh.
Dredging would be carried out at a rate of approximately 400 yd3
sediment/day working 8 hr/day. The dredged material would be
pumped ashore to the treatment facilities through a float-supported
pipeline. The pumping rate of the water-sediment dredge slurry
with approximately 20% solids by volume would be 1,000 gal/m.

Sediment Treatment and Disposal
  The dredged sediments would be treated at a facility constructed
on the former battery plant site. Figure 2 shows a schematic flow
diagram of the sediment treatment and disposal considered in the
remedial alternatives. The dredged sediment would first be gravity
thickened and then be either chemically fixated or  mechanically
dewatered.  The chemically fixated sediments  would  be non-
hazardous and  subsequently could be disposed of either off-site
in a proper municipal landfill or on-site in the landfill to be con-
structed on  the  former battery plant  site.  The  mechanically
dewatered sediments would be hazardous and would have to be
disposed of in a permitted RCRA landfill. The supernatant from
the gravity thickening would be clarified with the addition of a
coagulant and a coagulant aid and discharged to East Foundry
Cove Marsh and  Constitution Marsh.

Marsh Restoration
   After hydraulic dredging of any marsh areas, these areas would
be restored by replacement of sediment and replanting of vegeta-
tion. Sediment  would be replaced by  backfilling the marsh with
sand and topsoil to the original marsh elevation. For marsh revege-
tation,  the following concepts and  scheme were formulated:

• The disturbed marshes would be reestablished with narrow-leaved
   cattail (Tvoha  anoustifolia)  and arrow arum (Peltandra  vir-
  ginica). While greater marsh diversity would be desirable, there
   are relatively few species that would tolerate and thrive under
   the existing salinities  and flooding  regime
• Planting 2 year old plants would be recommended rather than
  seeding
• Plants would be installed on a grid pattern 2 ft center-to-center,
   alternating species so each species would be planted  on a  4 ft
   center-to-center grid
• Planting would be done from December  through March using
   dormant stock  or from April through mid-May using growing
  plants. Planting later in the  season results in too  much plant
  material to be  handled
•  Some of the planted sprigs  may not  survive due to normal
  planting mortality, and may have to be replaced annually for
  the first 2 to 3 years after initial planting

Containment
  To cap contaminated sediments, an Armorform mat was con-
sidered for its ease of installation (especially on very wet soils and
underwater) and for its venting capabilities.  Armorform mat has
been  widely used for erosion-control. It is a permeable, con-
tinuously woven panel of double-layer synthetic fabric jointed
together to form a framework for placing sand/cement mortar.
The permeable fabric restrains the loss of solids during underwater
mortar injection and allows excess mixing water to escape assuring
a low water/cement ratio, accelerating hardening and producing
a durable concrete structure. Installation of an Armorform mat
does not require dewatering of the site. Form work panels can be
positioned and mortar filled from the surface in shallow water.
                                                   FIXATED
                                                  SEDIMENTS
                                                (NON-HAZARDOUS)
                                         OFF-SITE/ON-SITE
                                             DISPOSAL
                          Figure 2
                Summary of Detailed Evaluation
Evaluation and Ranking
  After the detailed evaluation of each remedial alternative with
respect to each of the criterion listed in Table 6, the results were
expressed in a rating system utilizing the terms high, moderate and
low. A high rating indicates that the alternative promotes the intent
of the criteria and meets or exceeds the remedial objectives, for
the most part. A moderate rating indicates that the  alternative
neither promotes nor adversely affects the intent of the criterion.
With  a moderate rating, however, the alternative does remediate
the problem to an appreciable extent even though it does not meet
all the objectives. A low rating indicates that the alternative does
not promote the criterion and does not meet the remedial objectives
for the most  part.
  Table 7 summarizes the results of the detailed evaluation of the
remedial alternatives. The results of the non-cost evaluations are
summarized for each alternative using the high-moderate-low rating
                                                                                                 SITE REMEDIATION     345

-------
                            Table 7
                 Summary of Detailed Evaluation
                        lockitlcil  fmtromtil  ImtltutloMl Mile Null*   Coil
                        Fiillkllllj   l-p.ct.   lo~lro-.li lowlro.o.11  •"•ft"
                         ซปllm    MilBQ	SUM	SlSJjl	II "IH'ril
i. tilt fOU

  (CM-1

  ICh-l


  tCN-3


  ICH-4
        * Coปt Hปrปh (tin)
       No Action

       MjtnulK DrMllM/TMchmlM/
       Mullonratt-llll Dllpoul

       HJUnultc Orซ4ซIIM/nilGkm1lM/
       Muitป/0ii-stti Dlipoul
High


Xl*
                                                  t.n

                                                  re. II

                                                  it.M
   Ofl

   CH-1
HO Action


Mut Ion/Off-tfto'oflpoiol


OOMttrlMf/Off-SIU Dllpoul
                                                         c.n

                                                         ri.ii

                                                         n.i)
 • loul rroiml Urtl OHM m ป-^.r prajoct llfo II In 
-------
  contaminated (Cd, Ni and Co) sediments in East Foundry Cove
  Marsh can be developed by combining certain developed remedial
  technologies.  Bench-scale  tests have demonstrated the  site-
  specific application of these technologies.
• "No  Action" is  a viable remedial  alternative  for Consti-
  tution Marsh.

ACKNOWLEDGEMENT
  The work described in this presentation was funded by the U.S.
EPA  under U.S.  EPA  Contract No.  68-01-7250  with Ebasco
Services Incorporated. The contents do not necessarily reflect the
views and policies of the U.S. EPA, nor does mention of trade
names or commercial products constitute endorsement or recom-
mendation for use.

REFERENCES
1. Acres International Corporation, "Remedial Investigation at Marathon
   Battery Federal Superfund Site, Cold Spring, New York." A draft report
   submitted to New York State Department of Environmental Conser-
   vation, Aug 1985.
2.  Acres International Corporation, "Feasibility Study at Marathon Battery
   Federal Superfund  Site, Cold Spring, New York." A draft report
   *  submitted to New York State Department of Environmental Conser-
   vation, Aug 1985.
3.  Ebasco Services Incorporated, "Supplemental Remedial Investigation
   Report, Marathon Battery Company Site (Constitution Marsh and East
   Foundry Cove , Village of Cold Spring, Putnam County, New York."
   A final report submitted to U.S. EPA, Aug 1986.
4.  Ebasco Services Incorporated, "Supplemental Feasibility Study Report,
   Marathon Battery Company Site (Constitution Marsh and East Foundry
   Cove , Village of Cold Spring, Putnam County, New York.) A final
   report submitted to U.S. Environmental Protection Agency, Aug  1986.
5.  Renwick, W.H. and Ashley, G.M., "Sources,  Storages and Sinks of
   Fine-Grained Sediments in a Fluvial-Estuarine System." Geological Soc.
   of Amer. Bull. 95,  1984, 1343-1348.
                                                                                                    SITE REMEDIATION     347

-------
                          Hydrogeologic Assessment,  Delineation
                    And  Remediation Of a  Shallow  Groundwater
                                           Contaminated  Zone

                                                Martin M. Fontenot
                                            CIBA-GEIGY Corporation
                                               St. Gabriel, Louisiana
ABSTRACT
  This paper details a case history of a groundwater contamina-
tion episode and subsequent corrective action. A groundwater
monitoring system was installed in December 1981, in the vicinity
of several hazardous waste facilities. Groundwater samples col-
lected and analyzed from a well adjacent to a by-product HCL
storage tank during the second quarter of 1984, showed carbon
tetrachloride (CCl^ and chloroform (CHC13) concentrations in
excess of the method(s) detection limit of 0.01 mg/1, 11  ft. below
land surface (ft bis).  Resampling and subsequent analysis per-
formed confirmed the presence of these  two U.S. EPA-defined
priority pollutants in concentrations  of: (CHC1,—100 mg/l) and
(CC14—50 mg/1). A groundwater quality assessment program was
designed and implemented to determine the  rate and  extent of
migration of the two constituents. The groundwater quality assess-
ment program was designed to adequately define the vertical and
horizontal extent of the contaminant plume.
  Monitoring wells were installed in the vicinity of the tank in three
different zones, i.e. ฑ 8 ft. bis, ฑ 15  ft. bis and 40 ft. bis). A total
of 28 monitoring wells were installed, four of which were inclined
at 45ฐ to monitor groundwater quality beneath the tank.
  Water quality data from the wells showed the plume of contami-
nation to be confined to an area of approximately 10,000 ft.2. The
highest levels of CC14 and CMC 13 detected  were 5 mg/1 and 10
mg/1, respectively. The hydrogeological data was used to formu-
late a Corrective Action  Plan.
  The Corrective Action Plan involves removal of the tank, piping,
associated  equipment and contaminated  soil down to the water
table. The contaminated soil will be disposed  of off-site  in a com-
mercial hazardous waste landfill. The groundwater in the area will
be removed via a subsurface drain system and treated in the plant
wastewater treatment system prior to discharge. The excavated area
will be backfilled with  clean soils. Natural infiltration will induce
flushing of the contaminants to the drains. It  is estimated that the
cleanup period will be 1  year.

INTRODUCTION
  CIBA-GEIGY Corporation's St. Gabriel plant is located on the
east bank of the Mississippi River approximately 20 miles south
of Baton Rouge, Louisiana.  Specific processes at  the facility
include: (1) manufacture and formulation of s-triazine herbicides;
(2) manufacture of hydrogen cyanide, a raw material; (3) formu-
lation  and  packaging activities for various pesticides; (4) process
development activities; and (5) supportive activities for the above
which include effluent treatment systems, maintenance, utilities,
analytical and quality control.
  As a result of the above processes and services, non-recyclable
waste streams are generated. These waste streams were classified
as hazardous under the Louisiana Hazardous Waste Management
Plan (LHWMP) rules and regulations of 1979. The state's require-
ments for groundwater monitoring in the LHWMP.
  In compliance with the LHWMP, the St. Gabriel plant developed
and installed a groundwater monitoring system in  the latter part
of 1981. Some of the facilities identified as hazardous waste units
under the LHWMP included by-products HCL storage tanks.
  The groundwater monitoring system was sampled and analysis
was performed on a quarterly basis from December 1981 through
1984. The second quarter 1984 groundwater samples from a well
adjacent to one of the by-product  HCL  storage  tanks showed
carbon tetrachloride  (CC1J and chloroform (CCl,)  concentra-
tions in excess of the methods detection limit of 0.01 mg/1, 11 ft.
bis below land  surface.
  No other groundwater samples collected from the monitor well
network had previously indicted the presence of these two consti-
tuents. Resampling of the wells confirmed the presence of these
(volatile organic compounds) VOCs which are defined as U.S. EPA
priority pollutants. The results are presented in Table 1.
                         Table 1
     Concentration of VOCs in Groundwater Monitor Well (>ig/D
                      Dates Sampled
     VOC's
     Carbon Tetrachloride
     Chloroform
5/31

 69
                                81
6/27

 U9

 96
7/13

 31

 96
   In accordance with the Louisiana Hazardous Waste Regulations
 (Interim Status), the Department of Environmental Quality (DEQ)
 was provided written notification that two hazardous constituents
 were detected in a monitoring well in the vicinity of the by-product
 HCL storage tank.
   A groundwater consulting firm was contracted to develop and
 prepare a hydrogeologic assessment plan to comply with the State
 Regulations. A subsequent groundwater quality assessment pro-
 gram was implemented to define the contaminant plume vertically
 and horizontally. The results of the groundwater quality assess-
 ment program were used to develop a corrective action plan to
 delineate and remediate the contaminated shallow groundwater
 zone. The groundwater assessment program and the corrective
 action plan were approved by the State Environmental Agency;
 the corrective action plan is presently being implemented.
348    SITE REMEDIATION

-------
HYDROGEOLOGIC ASSESSMENT PLAN
  The assessment plan was designed to provide an evaluation and
determination of the rate and extent of the migration of CC14 and
CHC13.  The  ultimate objective  of the assessment plan was  to
generate sufficient data to facilitate the development of a Correc-
tive Action Plan to completely remediate the observed groundwater
contamination. The assessment was conducted in three phases:

• Phase I   - Preliminary Assessment.
• Phase II  - Additional  Data Collection.
• Phase III - Final Data  Collection and Recommendations for
  Corrective Action.

  Each  assessment  phase  is  described below.

Phase I
  The first  phase of the  Hydrogeologic Assessment Plan was
designed to provide the information necessary to delineate the
extent of migration  of CC14 and CHC13 in the vicinity of the con-
taminated well (9B). To accomplish this objective, a monitor well
network was designed consisting of four shallow wells (ฑ 8 ft. bis)
and 1 deeper well (ฑ 40 ft. bis) as shown in Figure  1. The net-
work was designed on  the assumptions that the contaminant
source(s) was close  to well 9B  and that the contaminant  had not
migrated vertically  or horizontally any significant distance. The
three shallow wells, 91,  92,  93  were installed  in a triangular
arrangement around well 9B  to  bracket the limits of the con-
taminants in the shallow aquifer.  Wells 9A and 9C were installed
near 9B to obtain a vertical profile of the groundwater levels and
quality  with  depths.
         Oป7
                         rxpum*Tion

                    81 •  DdSTMC MONITOR-WELL
                         LOCATION AND NUMBER

                    97 O  PHASE I MONITOR-WELL
                         LOCATION AND NUMBER
                           Figure 1
                Location of Phase I Monitor-Wells
                                                10 FEET
  The general hydrogeology in the vicinity of the acid tank was
defined by collecting continuous soil samples during the installation
of well 9C. The soil boring log for this well revealed the following
subsurface stratigraphy to be present:
     Stratum
     Silty CLAY
     CLAY  (A-Zone)
     Clayey  SILT (B-Zone)
     CLAY
     Clayey  SILT (C-Zone)
     CLAY
        Depth
Ft.  Below Land Surface
          0-2
          2-7
         7-12
         12-29
         29-34
      34-40 (min.)
  Three hydrogeologic zones were recognized and defined:

• A-Zone: The uppermost water bearing zone extending from 2
           to 7  ft bis
• B-Zone: A clayey silt layer extending from approximately 7 to
           12 ft bis
• C-Zone: A deeper clayey silt layer extending from 29 to 34 ft bis

  The surface stratigraphy is depicted graphically  in Figure 2.
                                                                       I-
                                                                                    m
                                                                                                   MMIWTAL KM1
                                                                                    JL I./I IOC Mid U
                                                                                              Figure 2
                                                                                   Hydrogeologic Cross-Section A-A'
Phase II
  The Phase II investigation consisted of the installation of eight
shallow monitoring wells in the vicinity of the acid storage tank.
The wells were screened from 3 to 8 ft. bis in the A-Zone (Figure
3). The monitor-well locations were selected to provide the infor-
mation necessary to delineate the lateral extent of the contaminant
plume in the vicinity of the acid storage  tank.

Phase III
  The Phase III investigation consisted of installing five A-Zone
wells (3  to 8 ft. bis), five B-Zone wells  (8-12 ft. bis) and one C-
Zone well (29-34  ft. bis).  Finally, four inclined boreholes were
drilled at an angle  beneath the acid tank. The location of the Phase
III monitoring wells is depicted in Figure 4. The objectives of the
Phase III assessment were to: (1) determine the extent and rate of
migration of groundwater contamination in the vicinity of the acid
tank and (2) make recommendations for a Corrective Action Plan.
  Twenty-eight wells were installed during the three phases of the
assessment.

GROUNDWATER FLOW
  Groundwater elevations were measured in all the assessment wells
over a 2-day period.  The measurements revealed that, hydro-
geologically, the A and B zones act as a single hydrogeologic unit;
that is, there is little or no difference in the behavior of ground-
                                                                                                 SITE REMEDIATION     349

-------
                 idoO
 On
              MO
             MO
                     • 17
                              91* DKWC HOMTOI-ซOi
                                  LOCA1KM AND MMEX

                              97 O nulC I I HOMTW-HEU
                                  UKA1XM AW MMER
                            Figure 3
                 Location of Phase II Monitor-Wells
                100 •
• n
            • 94
                    • 97
                                 OPLAIIATION
                             A-A'  OCOLOOCAL CROSS-ZCTKJH

                             tl*  DJSTWO MOMTOI-WU.
                                 LOCATION AW NUMO

                             97 O  PHA3C I I I VCMTOK-XU
                                 LOCATION AMI NUtlM*

                              •  MOWED PHASE I I I Mil
                            Figure 4
                Location of Phase III Monitor-Wells
water flow between these two units. A contour map showing the
elevation of the groundwater surface in this hydrogeologic unit is
provided in Figure  5.  Horizontal  groundwater  flow in  the
A-Zone/B-Zone was defined as being towards the south and west
under an average hydraulic gradient of 0.025 ft/ft. The average
linear horizontal groundwater velocity (V) in this zone was calcu-
lated using the formula:

v =  ^. i                                               o,
      n    L
Where:
Kh = Horizontal Hydraulic Conductivity
 n =  Porosity
-  = Hydraulic Gradient

   Using the  measured hydraulic gradient  (0.025), an assumed
representative value of 1  x  10 ' cm/sec for Kh and an assumed
value of 0.35 for porosity,  V was calculated  to be 7.1  x  10"'
cm/sec (2.0  x 10"' ft/day or less than 1 foot/year). A compari-
son of the groundwater evaluations in Zones A, B and C indicated
that a downward flow potential from the A-Zone/ B-Zone to the
C-Zone exists beneath the acid tank area.
                                                                                                          in O  c-zoe
                                                                                                               LOCATION ANOMMO
                                                                                                           •*— awuNO-wtoi fujn oacciot

                                                                                                               o   10 nn
                                                                                               Figure 5
                                                                                 A-Zone/B-Zone Water Level Elevations
GROUNDWATER QUALITY
  Groundwater  samples were collected from all the wells and
analyzed for  VOCs. The analyses showed that three parameters
(chloroform, carbon tetrachloride and toluene)  were present in
samples collected from the A-Zone/B-Zone monitor wells. Chloro-
form and toluene were detected in samples collected from 1 of the
C-zone wells. Concentration contours for chloroform and carbon
350    SITE REMEDIATION

-------
tetrachloride are shown for the A-Zone/B-Zone and for the C-
Zone in Figures 6, 7 and 8, respectively.
                  A-ZONE/B-ZONE WELL WITH
                  CHLOROFORM CONCENTRATION (ppm)

              9C O C-ZONE MONITOR-WELL
                  LOCATION AND NUMBER

               f INCLINED A-ZONE/B-ZONE WELL
                                                                      US.
                                                                       NO
                                                                                                      f5 • A-ZONE/B-ZONE WELL WITH
                                                                                                          CARBON CONCENTRATION (ppm)

                                                                                                      9C O C-ZONE MONITOR-WELL
                                                                                                          LOCATION AND NUMBER

                                                                                                       J INCLINED A-ZONE/B-ZONE WELL
                                                                                                   Figure 7
                                                                                            A-Zone/B-Zone Carbon
                                                                                          Tetrachloride Concentration
                             Figure 6
             A-Zone/B-Zone Chloroform Concentration
  The contour plots show that the lateral  movement of con-
taminants in the A-Zone/B-Zone was limited to an area northwest
of the acid tank. The concentrations of carbon tetrachloride and
chloroform in  the  majority of the samples collected  from the
A-Zone/B-Zone wells were at or below 1 (mg/1). Three wells in
A-Zone/B-Zone showed concentrations significantly greater than
1 mg/1 level. These include:

 Well      Chloroform (mg/1)     Carbon Tetrachloride (mg/1)
  93
 101
 DH1
9.8
9.5
2.7
 4.2
4.0
11.0
  These concentrations indicate areas of localized leakage of con-
taminants into the shallow groundwater system. Low concentra-
tions of toluene were measured in five wells. These include:
           Well
           DH1
           DH3
           11B
            93
           IOC
                   Toluene (mg/1)
                        0.02
                        0.015
                        0.035
                        0.56
                        0.062
  Both  carbon  tetrachloride  and  chloroform  have  similar
                                                        EXPLANATION

                                              un /„ ~ " ฐ  C-ZONE MONITOR-WELL WITH
                                              NO/0.014    CARBON TETRACHLORIOE
                                                        CHLOROFORM CONCENTRATION

                                                    Figure 8
                                    Tetrachloride/Chloroform Concentrations In
                                                  The C-Zone
10 FEET
                                                                                                       SITE REMEDIATION     351

-------
hydrogeochemical properties, i.e.:

• Both have a very low potential for adsorption to soils
• Both degrade very slowly
• Both are heavier than fresh water.

  The site hydrogeological properties together with the parameter
characteristics provide an explanation for the observed distribu-
tion of contaminants in the shallow groundwater system. The con-
taminants were released from the surface units, dispersed laterally
and then sank to the A-Zone/B-Zone. The contaminants have sub-
sequently been transported very slowly southward by the horizontal
groundwater  flow.

Contaminant Sources
  The suspected source(s) of the shallow  groundwater contami-
nation are:
• Acid Tank
• Pump Pad
• Tank Scrubber
  All of the above units are involved in the transfer, storage, treat-
ment  and  management of the acidic wastewater generated at the
St.  Gabriel Plant. The acidic wastewater  is neutralized and dis-
charged through the plant's NPDES wastewater treatment system.
This wastewater sometimes contained minor amounts of carbon
tetrachloride  and chloroform during process upsets, especially
during the early days of plant operation.  Therefore, any spilled
and/or leaked wastewater may eventually have caused contami-
nation of the shallow groundwater. The possible, sources of toluene
contamination may have been due to losses incurred during instal-
lation of the acid tank liner in which toluene was used as a solvent.
  The low concentrations of the constituents detected in the deep
well IOC do not appear to be a result of the wastewater manage-
ment  activities. We  believe that these constituents were inadver-
tently introduced into the C-Zone during  the installation of well
IOC.

Problem Identification
  The three phase assessment of the groundwater quality beneath
the  acid tank and associated facilities  resulted in the following
conclusions:
• The area is  underlain by alternating layers of clay and silt. The
  shallowest water-bearing zone  beneath  the site is a combined
  layer of clay and silt (A-Zone/B-Zone). Groundwater flow in the
  hydrogeologic unit is south and west beneath the site. There is a
  net downward flow potential from the shallow unit to a deeper
  waterbearing stratum (C-Zone).
• Limited contamination of groundwater with carbon tetrachlor-
  ide, chloroform and toluene in the A-Zone/B-Zone has occurred.
  The three facilities (Acid Tank, Pump Pad and HC1 Scrubber)
  have been identified as possible sources  of this contamination.
  No contamination was detected in the  C-Zone.
• Laterally, the groundwater contamination has  migrated north
  of the Acid Tank and HC1 Scrubber i.e. opposite to the direc-
  tion of groundwater flow. This is thought to be due to a com-
  bination of: (1) the localized leakage/spillage of fluids at the
  site; (2) chemical dispersion;  and  (3)  the low groundwater
  velocity  in  the A/B-Zone.
• The constituents detected in well IOC are  believed to be the result
  of  cross-contamination  between  the shallow and deep zones.
  These constituents appear to have been inadvertently introduced
  into the C-Zone during installation  of  well IOC.
elements:
• Defining the extent of subsurface contamination
• Outlining a course of action
• Implementing the course of action

Extent of Subsurface Contamination
  Figures 6 and 7 illustrate  the distribution of carbon tetrachloride
and chloroform plumes in  the shallow groundwater system. The
line of zero concentration shows the full extent of lateral ground-
water contamination The chloroform plume  is somewhat  larger
than the carbon tetrachloride plume and covers an area of approx-
imately 14,500 ft2.
  To determine the vertical extent of the groundwater contami-
nation, two hydrogeologic  cross-sections (Figures 9 and  10) were
constructed across the site. These show that  both  contaminants
have migrated to the base of the B-Zone. This phenomenon is
expected since both compounds have a density greater than fresh
water.
                           Figure 9
                 Chloroform Plume Cross-Section
                           Figure 10
             Carbon Tetrachloride Plume Cross-Section
 Corrective Action Plan
   The purpose of the Corrective Action Plan (CAP) was to
 develop, for approval by the DEQ, the best technically sound and
 cost-effective approach for removing the contaminated ground-
 water and its source from the site. The CAP involves three major
 CORRECTIVE PLAN OUTLINE
   The CAP is  being implemented in a three-stage process:
 •  Removal and decontamination of existing monitoring wells, Acid
   Tank, HO Scrubber, Pump Pad and all associated pipes, pumps
   and fixtures
 352    SITE REMEDIATION

-------
• Excavation and removal of soils that directly overlie the shallow
  water table within a specified area
• Installation of a subsurface drain system that will direct the con-
  taminated groundwater to a sump for removal to the plant
  wastewater treatment system

  The CAP is a remediation approach, whereby the contaminant
sources and the unsaturated soil zone will be removed. The con-
taminated groundwater will be removed by continuous purging
through a subsurface drain system and subsequently will be treated
in the plant wastewater treatment facilities.

Equipment Removal
  The storage tank and associated equipment will be decontami-
nated by rinsing/ flushing with high pressure water. The washwater
will be treated in the plant wastewater treatment facilities. The tank
and associated equipment will be salvaged if possible; otherwise,
they will be disposed off-site  in an  approved hazardous waste
landfill.
  The monitoring wells installed during the hydrogeologic assess-
ment will be plugged and abandoned design to prevent the down-
ward vertical migration of contaminants. The well casings will be
removed, if possible,  prior to  filling the well  with  a cement-
bentonite grout. Four of the wells will be retained as  post CAP
monitoring wells (91, 92, 99 and 105, in Figure 4).

Soil Excavation
  The unsaturated soils within the area shown in Figure 11 will
be excavated to a depth just above the water table (approximately
3 ft. bis). This represents a total of 1,750 yd3 of potentially con-
taminated soil that will be removed  for off-site disposal.
                                  4'-INCH DIAMETER FEEDER LINES
                                  INSTALLED ON A GRADE Of 0.1X
               I	1
               I     PUMP    I
               !     PAD
               I	1  /
                            /
         f-\MCH DIAMETER HEADER    /
         INSTALLED ON A GRADE Of 0.2X /
           I    HCL
             SCRUBBER
           I	1
                                                                                                              • DIRECTION OF FLOW

                                                                                                               0   10 FEET
                            Figure 11
                   Proposed Area of Excavation
                                                                                              Figure 12
                                                                                  Proposed Subsurface Drainage System
Subsurface Drain System
  The subsurface drainage system will be installed at the base of
the permeable B-Zone. The depth of the B-Zone is approximately
12 ft. bis and 9 ft. below the initial excavation. The subsurface
drainage system consists of 4-in. diameter, perforated corrugated
plastic tubing feeder lines which will collect the contaminated
groundwater  and transport it to a common 6-inch diameter un-
perforated header pipe as illustrated in Figure 12. The header will
transport the water to a sump for removal to the plant wastewater
treatment system. The plant wastewater treatment system consists
of a multi-train activated carbon adsorption system, which is well-
suited for the removal of organic contaminants from wastewater.

POST-INSTALLATION MONITORING
  A monitoring program will be implemented to determine the rate
at which the drainage system is removing  the contaminated ground-
water. The program will consist of collecting and analyzing monthly
samples from the four groundwater monitoring wells (91,  92, 99
and 105 in Figure 4) and the sump discharge. These wells will be
monitored for the duration of the drain system operation. In ad-
dition, monthly water samples will be collected from the sump dis-
charge line and analyzed for carbon tetrachloride and chloroform.
When the concentrations of these constituents are below the de-
tection limit, the site will be deemed free of groundwater contami-
nation. The  time  required for site decontamination  will be  a
function of:

•  Rate groundwater is removed from the system
• Amount of contaminant retardation
• Mass of contaminants in the system
                                                                                                  SITE REMEDIATION    353

-------
  Recharged water (natural or induced) will act to flush the sur-
face drains. We estimate one year will be required to reduce the
maximum observed contaminant concentration in the groundwater
to the established background groundwater quality (below the
detectable limit of 0.01 mg/1).
  At the end of the groundwater cleanup, all of the drain piping,
sump and monitoring wells will  be removed in accordance  with
standard agency protocols.

AGENCY APPROVAL AND STATUS
  The DEQ was provided notification of the detected groundwater
contamination in accordance with the Interim Status Hazardous
Waste Regulations.  Subsequently,  a Hydrogeologic Assessment
Plan was submitted to and a meeting was held with the DEQ to
discuss the results of the assessment and proposed CAP. The DEQ
approved the CAP in August 1985.
  The acid storage tank facility is being replaced with a new facil-
ity. This new facility will meet standards applicable to hazardous
waste storage tanks, although the by-product HCL is not RCRA-
regulated. The new facility is scheduled to be completed by Novem-
ber 1986. The implementation of the CAP is scheduled in early
 1987.

 ACKNOWLEDGEMENTS
  The author gratefully acknowledges the assistance of the Baton
 Rouge, Louisiana Office of Geraghty & Miller, Inc., Groundwater
 Consultants.

REFERENCES
 1.  Geraghty and Miller, Inc., "Preliminary Groundwater Quality
   Assessment,"  prepared for CIBA-GEIGY Corporation, St.
   Gabriel, LA.,  1984
2.  Geraghty and Miller, Inc. "Phase II  Status Report, Ground-
   water Quality Assessment," prepared for CIBA-GEIGY Cor-
   poration, St. Gabriel, LA., 1985
3.  Geraghty and  Miller, Inc., "Phase III Groundwater Quality
   Assessment,"  prepared for CIBA-GEIGY Corporation, St.
   Gabriel, LA.,  1985
4.  Geraghty and Miller, Inc., "Corrective Action  Plan for Acid
   Tank Site," prepared  for CIBA-GEIGY  Corporation,  St.
   Gabriel, LA.,  1985
 354    SITE REMEDIATION

-------
                      The  South  Valley  San  Jose  6  Superfund Site
                                    Special  Issues  and Problems

                                                   Kathleen O'Reilly
                                                 U.S. EPA Region  VI
                                                     Dallas, Texas
                                                     Steve  Tarlton
                                                     CH2M HILL
                                                   Denver,  Colorado
                                                       Paul Karas
                               New Mexico Environmental Improvement Division
                                                Santa Fe, New Mexico
 ABSTRACT
  Each Superfund site is unique in terms of physical and chemi-
 cal properties and political issues. The South Valley San Jose 6
 Superfund site in Albuquerque, New Mexico, has a number of
 special issues and problems from which investigators of other
 Superfund sites may gain knowledge.
  The San Jose 6 Superfund site is located in an industrialized area
 of southern Albuquerque. Municipal Well San Jose 6 is located
 in the center of this area. The lack of a fenced, well-defined site
 presented problems during well installation activities both in terms
 of security and health and safety protocols.
  The regional geology made the selection of drilling techniques
 an issue at this site. The dual wall drilling technique was used to
 achieve desired well completion depths and prevent vertical cross
 contamination in the borehole. It also minimized and facilitated
 waste generation and disposal during drilling.
  Waste disposal during drilling and sampling involved coordina-
 tion  with  federal, state  and local authorities. The authorities
 established disposal protocols for wastes of different contaminant
 levels by using an on-site mobile laboratory.
  A  special laboratory program was established for known car-
 cinogens, and 10~6 lifetime excess cancer risk concentrations were
 used as minimum  detection limits.  Laboratory  protocol was
 established as modifications to U.S.  EPA Method 624 to obtain
 quality data.
  In addition to the solvents present,  petroleum product contami-
 nation complicates regulatory jurisdiction at  the site. Solvents and
 petroleum contamination are handled separately under both
 CERCLA and RCRA. Instances of petroleum leaks from under-
 ground storage tanks, above ground storage tanks and pipelines
 are each regulated separately.

 INTRODUCTION
  During 1986 and 1987, RI studies  were completed on the San
 Jose  6 Superfund site in the South Valley area of Albuquerque,
 New Mexico. The planning and completion of these studies had
 to account for a variety  of special issues related to the project
 setting, monitoring well drilling, investigation waste management,
 laboratory analyses, regulatory jurisdiction and potential remedial
 actions.

Background
  The San Jose 6 Superfund site was placed on the NPL in 1981,
and U.S. EPA field studies at the site were initiated in 1984 in
accordance with CERCLA. The discovery of contaminants in
municipal well San Jose 6 prompted  the City of Albuquerque to
 take it out of service in 1980. An initial remedial measure (IRM)
 was initiated by the U.S. EPA to provide a replacement well,
 presently under construction.
   The site is located in the South Valley area of Albuquerque,
 Bernalillo County, New Mexico. The  site,  which includes both
 industrial and residential areas, encompasses approximately 2
 miles2.  It has been designated as the state's  highest priority
 Superfund site because hazardous substances are present in the
 groundwater near the city's San Jose well field.
   RI activities (Phase I) were conducted by the U.S. EPA and the
 PRPs from  1983 through 1985. In 1986, additional RI studies
 (Phase II) were initiated to provide the remaining necessary data.

 SETTING
   The setting of this project has complicated several aspects of
 the site studies. The site consists of large undeveloped tracts,
 numerous industrial properties and a few residences.
   The original studies were conducted by the New Mexico Environ-
 mental Improvement Division (NMEID), by each PRP within their
 property boundaries and by the U.S. EPA in the remaining areas.
 Thus, the data base for the site has been developed from more than
 seven studies conducted at different times.  In some cases, field
 procedures and analyses performed were inconsistent with each
 other, complicating data evaluation.
   Until 1986, groundwater levels had not been measured in the
 same time period in wells across the site. Groundwater levels could
 be mapped;  however, uncertainties were recognized because of
 temporal changes in groundwater levels.
   Another complication was that most PRP and other  industrial
 properties in the area are fenced for security, while other areas are
 unfenced and access is not physically restricted. Security and safety
 of site work in these open areas was a constant problem. Tem-
 porary barriers  and warning signs were installed but were only
 marginally effective. Fortunately, population and traffic in the area
 are light.

 WASTE MANAGEMENT
   The management of investigation  derived materials on  any
 Superfund site is a function of the setting, nature, degree and extent
 of contamination, regulatory constraints and investigation methods
 employed. Management protocols for investigation derived wastes
 during the most recent U.S. EPA studies were developed through
 discussions and negotiations between the U.S. EPA and its con-
 tractors, the NMEID and the  City of Albuquerque. The waste
 management plan included full containment of all investigation
.derived wastes, determination of degree of contamination  and
                                                                                            SITE REMEDIATION    355

-------
disposal appropriate for the contamination present.
  Contamination exists at depth in groundwater and soils but not
necessarily at the surface of the site. Surface dumping of drill
cuttings or fluids could  potentially have expanded the contami-
nated areas and created new exposure routes. The need to fully
contain drill cuttings and fluids affected the selection of drilling
techniques, as discussed in  a later section.
  The degree  of contamination was determined  predictively
through the use of previously obtained data and directly through
field analysis in a Close Support Laboratory (CSL). Cuttings and
liquids from new wells were analyzed by chromatography analy-
sis for 10 specific indicator compounds in an on-site mobile labora-
tory.  Disposal methods for water purged from existing wells were
determined by  reviewing previous water quality data.

DRILLING TECHNIQUE
  The selection of a drilling method for monitoring well construc-
tion depended upon cost, field time, nature of the contamination,
geologic conditions and drilling waste containment requirements.
The earlier studies had revealed that the alluvial of the aquifers
of concern are stratified  and loosely consolidated and that volatile
organics were present in groundwater. Borings for new monitoring
wells  were constructed to depths of more than 100 ft in areas where
the unsaturated zone was approximately 80 ft thick. Also, 250-ft
borings would  be constructed through the highly contaminated
flood plain aquifer into the possibly uncontaminated deeper zones.
  Methods employed recently for installation of monitoring wells
at the site involved both mud rotary and hollow stem auger drilling
techniques. The following  conditions led to the evaluation and
selection of an alternative drilling technique:

• Previous drilling programs conducted at the site had encoun-
  tered difficulties using mud rotary and hollow stem auger tech-
  niques.  Because of the  loosely consolidated sediments and
  associated high permeabilities, holes that were drilled using mud
  rotary techniques tended to slough, collapse or lose circulation.
  This led to well completion  above targeted depths, as well as
  significantly  increased time and costs for drilling and well com-
  pletion.  Holes that  were drilled using hollow stem auger tech-
  niques were limited in achievable depths and tended to experience
  refusal in shallow cobble beds.
• To be  able to detect the lower levels of contamination, down-
  hole introduction of materials had to be minimized and cross-
  flow between aquifers had to be prevented. Crossflow was of
  particular concern  since the shallow,  intermediate and deep
  aquifers are  progressively less contaminated and have strong
  downward vertical gradients. Conventional rotary or auger tech-
  niques would have required expensive telescoping of casing
  through each hydrologic unit. Rotary techniques also would have
  required the  addition of large quantities of drilling additives to
  maintain circulation and hole integrity.
• All liquids and solids generated during the drilling had to be
  tested for level of contamination and, depending upon the level,
  disposed of appropriately. Muds used in rotary techniques would
  have been expensive and difficult to dispose of at either a land-
  fill or  a  hazardous waste site because of their liquid composi-
  tion. Well development also would generate large volumes of
  water  if mud rotary  drilling were  used. Materials  generated
  using hollow stem augers would have been difficult to handle
  and contain.

  Based upon  the above mentioned conditions and the specific
objective of the investigation, the following criteria were developed
for selection of a drilling method:

•  Include  a drive casing to minimize crossflow between aquifers
  during well completion and minimize problems associated with
  maintaining  integrity of  the well  bore
• Minimize addition of materials downhole that may compromise
   water  quality data
• Minimize liquids and solids produced by the drilling since they
  must be contained and removed from the site. These wastes could
  potentially require costly disposal at a hazardous waste facility
• Provide representative  samples for  lithologic logging
• Allow for well completion in a manner that provides proper
  placement of well screen, gravel pack, bentonite seal and grout
• Well drilling and completion must be accomplished in a reason-
  able amount of time,  thereby limiting costs associated with
  drilling supervision  and meeting project schedules

  It was determined that the above objectives could best be met
  using either air rotary casing hammer or reverse air circulation,
  dual-wall, hammer  drilling methods.

ANALYTICAL ISSUES
  During the planning phase of this project, previous analytical
  results were  reviewed to identify potential levels of contamina-
  tion that  could be used to guide remedial actions. Remediation
  guidelines often used for carcinogen and suspected carcinogen
  contamination are the 10"* cancer risk concentrations.  Routine
  analytical detection  limits  for the U.S. EPA CLP significantly
  exceed the  10 ~6 cancer  risk  concentration  for many  com-
  pounds. To  determine the extent of site contamination and/or
  to determine the quality of background locations, detection limits
  must  be lowered to determine whether or not a particular loca-
  tion is contaminated.
  An attempt was made via the Special Analytical Services (SAS)
of the CLP to use more sensitive analytical techniques. A suitable
U.S. EPA analytical protocol (proposed Method 524.1) was iden-
tified by the U.S. EPA; however, no laboratories within the CLP
were interested in performing these analyses for the San Jose 6 site.
  Other alternatives to this approach were explored and  resulted
in the following alternatives:

• Perform analyses using U.S. EPA Method 624. This  alterna-
  tive would not yield analytical results that would be sensitive
  in the "decision range" (10~6 cancer  risk concentration).
  Analyses could be performed  by the CLP program readily as
  per an established U.S. EPA method.
• Perform analyses using modified U.S. EPA Method 624. This
  method would achieve the desired sensitivities. Analyses would
  not be as per  an established U.S. EPA method, which would
  make court  defensibility of  results more difficult.
• Use proposed Method 524.1.  If a laboratory could be found that
  would use proposed Method 524.1, desired sensitivities would
  be achieved and approved U.S. EPA methods would be used for
  the analyses. Some  delays in field effort could have been ex-
  perienced while a qualified laboratory was located. Questions
  on laboratory qualifications may have arisen since such a labora-
  tory may not have  been subject to thorough CLP screening
  procedures.
• Use approved, appropriately sensitive U.S.  EPA GC Methods
  601 and 602  for analyses. It should be possible to achieve desired
  sensitivities with these established U.S. EPA analytical methods.
  The non-confirmatory nature of this GC analysis (versus GC/MS
  analyses used in Methods 524.1 and 624) would increase uncer-
  tainty about the interpretation of analytical results. The job was
  bid through  the SAS component of the CLP, but there were no
  respondent laboratories.

  The review of these  options resulted in the selection of the first
alternative, as discussed below:
  All laboratory analytical  samples are in accordance with the
routine analytical services (RAS) procedures established by the U.S.
EPA in the CLP program. The detection limits tor these analyses
are standard for RAS methods. Groundwater samples were ana-
lyzed for nine volatile organics using modified the U.S. EPA
Method 624 to provide lower detection levels (Table 1). This spe-
cial testing provided the precision necessary to evaluate contami-
nation at the site and the associated cancer risk.
 356    SITE REMEDIATION

-------
                           Table 1
       Volatile Organic Compound Analytical Detection Limits
                        San Jose 6 Site
Routine
Detection
Limits
per liter)
5
5
5
5
5
5
5
5
10
10"" Life-
Cancer Risk
Concentrations
per liter)
0.67
0.42
0.19
0.94
0.17
0.6
2.8
0.8B
1.0
Special
Detection
Limits
per liter)
0.21
0.13
0.20
0.35
0.2
0.5
0.18
0.5
0.5
  Volatile Fraction/Compound

  Benzene
  Carbon tetrachloride
  Chloroform
  1,2-dichloroethane
  1,1,2,2-tetrachloroethane
  1,1,2-trichloroethane
  Trichlorethylene
  Tetrachloroethylene
  Vinyl chloride
  To test this method, a comparison of the RAS method and the
modified method were performed during an initial sampling round.
Seven replicate samples were collected at two wells for both RAS
and modified analysis. A confidence limit of greater than 904 was
calculated for the modified method for the  five  compounds
detected,  while the RAS method failed to detect four of the five
compounds.
  The use of both methods was adopted to provide the required
detection limit and court-defensible data.

REGULATORY ISSUES
  The contamination identified at the San Jose 6 Superfund  site
fall under several different legal jurisdictions. The primary regula-
tory jurisdictions include  CERCLA (and SARA),  RCRA and
underground storage tank (UST), discussed below.

CERCLA (and SARA)
  Contaminants that fall under CERCLA jurisdiction at the San
Jose 6 Superfund site are primarily volatile organic compounds.
CERCLA covers an extensive list of contaminants, but excludes
petroleum, including crude oil or any fraction thereof. U.S. EPA
Region 6 manages the CERCLA program in New Mexico; however,
NMEID concurrence and participation are encouraged. State par-
ticipation is  required for fund-financed remedial actions.

RCRA
  RCRA applies to the management of hazardous wastes within
an active  facility and regulates the generation, transportation and
disposal of hazardous wastes. This law does not apply to products
or raw materials and excludes wastes associated with the explora-
tion, development or production of crude oil. RCRA has correc-
tive action requirements for accidents, spills or leaks similar to
Superfund remedial actions. In general, petroleum products do not
classify as hazardous wastes when released or spilled. U.S. EPA
Region 6 has delegated RCRA primacy in  New Mexico to the
NMEID, except for portions of RCRA amended in 1984 that still
are administered by the U.S. EPA.

UST
  Regulations applying to underground storage tanks are included
in separate sections of CERCLA and RCRA. RCRA provisions
regulate these  tanks and set procedures for corrective actions.
CERCLA, as amended by SARA, provides limited funds for cor-
rective actions from UST leaks in cases where a financially respon-
sible party is not available. UST regulations  are evolving and are
delegated by the U.S. EPA to the NMEID. These regulations
specifically exclude transportation pipeline facilities but include
petroleum products.

SUMMARY
  The groundwater contamination across the site is divided as to
regulatory jurisdiction. Petroleum-related compound contamina-
tion of the flood plain aquifer cannot be specifically addressed in
the RI or feasibility study unless it will  have an impact on a
CERCLA remedial action.
  Currently, the NMEID is conducting UST investigations at two
PRP properties. Agencies that  regulate pipelines are being noti-
fied of the suspected leaks in the petroleum pipelines south of the
site.

CONCLUSION
  A variety of special issues that complicated the work at the San
Jose 6 Superfund site were recognized and addressed in the planning
stages of the remedial investigation.  These issues included the
following:

• Because of the absence of physical boundaries, the project setting
  presented difficulties for  security and health and safety.
• Analysis of investigation-derived waste in an on-site mobile
  laboratory allowed prompt determination of appropriate disposal
  methods.
• An unusual drilling technique, the reverse air, dual wall, casing
  hammer  method was used to prevent cross contamination of
  different aquifers, and to minimize the production of wastes.
  Accelerated  drilling rates reduced field time and allowed the
  higher cost-per-foot method  to remain cost-effective.
• A modification of U.S. EPA Method 624  was used to provide
lower  detection limits  required  to estimate the  risk  from
  exposure at the  site,  in  conjunction with regular analytical
  methods to provide court-defensible data.
• The contamination  found at the site falls under a variety of
  regulatory jurisdictions including CERCLA, RCRA, UST and
  others, thus complicating the definition of the CERCLA problem
  at the site and affecting potential remedial actions.
                                                                                                SITE REMEDIATION     357

-------
                       Installation  of  Monitoring  Wells Using  the
                          Dual Wall Hammer  Drilling  Technique
                                                    Thomas  C.  Sale
                                                    Sara E.  Rhoades
                                                      CH2M HILL
                                                   Denver, Colorado
ABSTRACT
  The dual wall hammer drilling technique offers an effective
approach to the installation of monitoring wells in unconsolidated
sediments. For the monitoring well installation program at the
San Jose 6 Superfund site, program objectives were successfully
achieved using the dual wall  hammer drilling technique in an
environment where conventional approaches to monitoring well
installation would have been inappropriate.
  The dual wall hammer drilling technique involves dual wall drill
pipe, an open drive drill bit, reverse circulation of air and a pile
driver for advancing the drill pipe. The following are advantages
to using the technique:

• Use of a drive casing that limits contaminant migration in the
  borehole during drilling and well completion and limits problems
  associated with maintaining the borehole in noncohesive uncon-
  solidated sediments
• Minimal requirements for downhole drilling additives that might
  bias water quality samples
• Ability to drill various unconsolidated materials ranging from
  cobbles and gravels  to sands, silts and clays
• Rapid and cost-effective installation of 4-in. diameter monitoring
  wells in unconsolidated sediments to depths of and possibly
  exceeding 250 ft
• Quality lithologic samples allowing for  accurate characteri-
  zation  of subsurface stratigraphy

  The major  disadvantages  of the method are:

• The potential to produce drill cuttings in excess of the borehole
  volume
• Health and safety concerns  related to circulating air
• The potential to introduce oils through the air compressor.

  Although these were concerns,  they did not  present major
problems during the  San  Jose 6  monitoring well  installation
program.

INTRODUCTION
  The dual wall hammer drilling technique was used at the San
Jose 6 Superfund site, Albuquerque, New Mexico, to drill and com-
plete 19 groundwater monitoring wells. This paper describes the
dual wall hammer drilling method and discusses its advantages and
disadvantages. Specifically  presented  are  the criteria used in
selecting the  method,  a description  of the method,  the results
obtained and conclusions regarding future applications.
  The San Jose 6 site lies within the Rio Grande Rift and is under-
lain by a thick sequence of unconsolidated sediments. The site,
approximately 1.5  mi2  in area,  contains alluvial  and colluvial
deposits. These deposits predominantly consist of sands and gravels
but also include beds of cobble, silts and clays and thin alternat-
ing sequences of cemented and non-cemented fine sands.
  Various volatile and semi-volatile compounds have been identi-
fied in the soils and groundwater at the site. The presence of these
compounds resulted in shutting down the San Jose 6 municipal
well in 1980. This shutdown reduced the city's capacity to provide
water for municipal supply and fire protection. In Albuquerque,
groundwater contamination is of particular concern as the city relies
solely on groundwater for its water supply.
  The objective of the  San Jose 6 monitoring well installation
program was to characterize site conditions sufficiently to evaluate
technically feasible and cost-effective alternatives for site remedia-
tion. Requirements for site characterization included evaluating
the nature and extent of contamination, delineating the extent of
aquifers and aquitards and estimating aquifer hydraulic charac-
teristics.

Selection of Well Installation Technique
  Selecting an appropriate technique for a monitoring well instal-
lation program is a complex process of weighing program objectives
against constraints of site-specific conditions, costs, scheduling and
equipment availability. The following are specific project objectives
and constraints critical to selecting the monitoring well installation
program  for the San Jose 6 site:

• Representative Detection of Low Levels of Volatile Organic
  Compounds. To detect the anticipated low levels of contamina-
  tion that are near detection limits, introduction of drilling muds
  or other downhole additives and crossflow between aquifers had
  to be minimized.  Crossflow during drilling and well completion
  was of particular concern since the  shallow, intermediate and
  deep aquifers are progressively less contaminated and have strong
  downward vertical gradients.
• Containment of All Liquids and Solids. All liquids and  solids
  generated during the drilling had to be contained, tested for level
  of contamination and, depending on the level, disposed of in
  a  local landfill  (solids),  a  municipal sewer (liquids) or an
  approved hazardous waste disposal facility.
• Difficult Drilling Conditions. The target depths for this drilling
  program were between 30 and 250 ft below grade. Previous
  drilling programs conducted at the site had encountered diffi-
  culties using mud rotary and hollow stem auger techniques. Bore-
  holes drilled using mud rotary techniques tended to sluff, collapse
  or lose circulation, which led to well completion above targeted
  depths, as  well as significantly increased time  and costs for
358    SITE REMEDIATION

-------
  drilling and well completion. Boreholes drilled using hollow stem
  auger techniques were limited in achievable depths and tended
  to reach refusal in the shallow cobble beds.

Definition of Criteria
  Based on the previous objectives and constraints and the general
objectives of  the monitoring  well installation program, the
following  criteria were developed to select a drilling method:

• Prevent crossflow between aquifers during well installation
• Limit problems associated with maintaining the boreholes during
  drilling  and well completion
• Minimize  adding materials downhole that might compromise
  water quality data
• Minimize and contain liquids and solids produced by the drilling
  since they would be disposed of off-site and potentially require
  costly disposal at a hazardous waste facility
• Drill and  complete monitoring wells to depths of 250 ft.
• Provide representative drill cuttings for characterization of site
  stratigraphy through beds of cobbles, gravels,  sands, silts, and
  clay
• Allow for well completion so  that it provides proper placement
  of well  screen, gravel pack, bentonite seal and grout
• Minimize exposure of the field team to contaminants encountered
  during well installation
• Accomplish well drilling and  completion in a reasonable  time,
  thereby limiting costs associated with drilling  supervision and
  meeting project schedules

CONVENTIONAL TECHNIQUES OF WELL INSTALLATION
  Initially, rotary, auger and cable tool techniques were considered
for installing the monitoring wells. Although these approaches had
been used previously at the site, none could reasonably meet project
objectives within the dictated constraints.
  Air rotary techniques, without a drive casing, were discounted
due to the unconsolidated  drilling environment. Mud rotary tech-
niques were discounted due to the cross contamination potentially
associated with mud circulation, the need to telescope well  com-
pletions through each unit of lower hydrostatic head, the poten-
tial for lost circulation zones and the cost associated with the
disposal of drilling muds.
  Auger techniques were  discounted based on the need to  com-
plete wells to depths of 250 ft and their history of reaching refusal
in cobble  beds at the site as shallow as 40 ft below grade.
  Cable tool techniques came closest to meeting  the needs of the
monitoring well  installation program.  The major limitations to
using cable tools were  the slow rate of well completion and the
costs associated  with a lengthy field effort.

SELECTED ALTERNATIVES
  Since conventional techniques for the installation of monitoring
wells were deemed inappropriate, consideration was given to less
commonly used alternative methods. Based on a review of  alter-
natives, it was decided that the project objectives could be achieved
best using air rotary/casing hammer techniques or dual wall
hammer drilling techniques.
  Drill specifications were developed and sent out to bid for both
techniques. Layne Environmental Services, Inc., Phoenix, Arizona,
was awarded the contract using the dual wall hammer method.
Costs associated with using the air rotary/casing hammer technique
were significantly higher.

Method Description
  The dual wall hammer drilling technique was developed approxi-
mately 30 years ago for  mineral exploration. In the past 5 to
10 years, the method has been used for exploration and hazardous
waste drilling projects  as well as geotechnical investigations and
various construction applications. The rig is versatile and can be
converted to dual wall rotary drive,  drop hammer and diamond
drilling if subsurface conditions dictate a change in drilling tech-
nique. Switching between methods takes between 1 and 2 hr. These
options were not used during the San Jose 6 well installation pro-
gram but are worth noting. At the present time, there are only 10
dual wall rigs  in the United States.
  The drill rig used by Layne Environmental at the San Jose 6 site
was an AP-1000 manufactured by Drill Systems, Calgary, Alberta.
The method involves using dual wall drill pipe, reverse air circula-
tion  to carry cuttings to the surface, an open drive drill bit, and
a diesel pile driver to advance the drill pipe. Figure 1  presents a
diagram of the dual wall hammer rig used at the San Jose 6 site.
      AIR COMPRESSOR
                   HYDRAULIC DRILL PIPE PULLER •


                           Figure 1
                  Dual Wall Hammer Drill Rig
              H COMPMtlOH
                           Figure 2
             Circulation Path of Air and Drill Cuttings
                                                                                                   SITE REMEDIATION    359

-------
Drilling
  A rig-mounted air compressor (250 lb/in.2, 750 ftVmin) was
used to circulate air downhole through the outer wall and up
through the inner wall of the drill pipe. The produced cuttings and
water were carried to the surface through the inner wall of the drill
pipe and discharged through a flexible hose and a cyclone to 20-
to 30-yd3 lined steel dumpsters. The cyclone reduced the velocity
from the  discharge and served as the sample collection point.
Figure 2 details the flow pattern of air and cuttings through the
drill pipe  and cyclone.
  The drill pipe was advanced by driving the pipe into the ground
using a diesel pile driver capable of exerting up to 8,100 ft-lb/blow
at rates up to 95 blows/min. Average drilling speed  from initiation
of drilling to total hole depth was approximately 100 ft/hr.

Sampling
   Grab samples were retrieved as the  cuttings were discharged
through the cyclone and into the containment bin. The samples
were collected for lithologic logging and analytical screening to aid
in selecting an appropriate disposal site for  the drill cuttings and
liquids.
   The grab sampler was constructed by attaching a fine wire mesh
sieve to a 5-ft-long, 1-in. diameter wooden dowel. The 5-ft handle
enabled the individual taking the samples to remain a sufficient
distance from the discharge point for health and safety purposes.
   Composite samples were recovered in 5-ft  intervals or at signifi-
cant changes in lithology.  In addition to the composite samples,
relatively  undisturbed clayey soil  samples were obtained  for
chemical analysis and permeability tests by driving a Shelby tube
sampler.
   The drilling method provided  immediate and  representative
formation samples to the discharge point due to the rapid penetra-
tion of the drill and high volume of air circulated through the drill
pipe. This technique provided a fairly continuous and accurate
stratigraphic log of the borehole throughout the drilling process
and allowed for accurate  selection of well screen  depths. Minor
difficulties did occur in defining exact  depths to lithologic units
when drilling through clayey material  since the clays  tended to
intermittently and temporarily clog the discharge line.

Completion
   The dual wall drill pipe is furnished in 10-ft lengths and has a
9-in. outside diameter and  a 6-in. inside diameter. The 6-in. inside
diameter allowed a maximum well diameter completion, inside the
drill pipe, of 4 in. using an artificial gravel  pack. After reaching
the borehole's total depth, air was circulated until the water being
produced  ran clear and the drill pipe was free of solids. At this
point, the air was shut off and the well screen and casing were run
down the  inner 6-in. inside diameter opening of the drill pipe and
the open drill bit. All casing and screen were steam cleaned prior
to installation.
  The gravel pack was installed through  the annular space between
the 6-in. inner wall of the drill pipe and the well casing. Sufficient
gravel was placed in the drill pipe to place an envelope around
10 ft of well screen. Next, a 10-ft joint of drill pipe was pulled.
The drill  pipe  was withdrawn using  a hydraulic  pipe puller,
operating with a 100-ton pulling force. Maintaining the gravel pack
levels consistently inside the drill pipe, gravel pack was added, and
joints were sequentially pulled until the drill bit was approximately
1 ft above the desired level for the gravel pack.
   At this point, air was again circulated downhole to remove excess
gravel pack from the drill pipe. If  only small amounts of gravel
pack were carried to the surface, additional gravel pack was added
to the hole and the hole was blown again to test whether the gravel
envelope had been placed to the desired level around the well screen.
Finally, a small volume of fine sand was placed at the top of the
gravel pack to further inhibit grout intrusion into the gravel pack.
   In some instances, due to unstable formations or shallow com-
pletion depth, the gravel pack elevations could not be determined
by blowing air through the drill pipe. For shallow well completions,
gravel pack  levels were sounded with a weighted tape. For deep
completions in unstable formations, the gravel pack could not be
accurately tagged so levels were estimated using the observed and
calculated fill rate of 2 ft per 100 Ib of gravel.
  After placing the gravel pack, a 1-in. diameter flush joint tremie
pipe was run down the 6-in. inner wall  of the drill pipe alongside
the well casing. This pipe was used to place a 2-ft bentonite seal
on top of the fine sand to further protect against grout intrusion
into the gravel  pack. After the seal was placed, the tremie pipe
was removed, and the seal was allowed to hydrate for 30 minutes.
  Next, the space between the inner wall of the drill pipe and the
well casing  was filled  with grout.  For grout  placement to the
surface, the annular space between the inner wall of the drill pipe
and the well  casing was used as the tremie pipe. The remaining
steps to completing the well involved pulling the remaining  10-ft
dual wall drill pipe out of the hole while keeping the drill pipe full
of grout. A  continuous grout seal was obtained by maintaining
grout in the pipe.

Drilling and Completion Variation
  The above described method of drilling and completion was used
in most of the 19 wells that were installed. Loose fine-grained sands,
which tended to flow, were encountered in several holes below the
water table.  Due to these conditions, variations were required to
the drilling and completion procedures in deeper wells.
  The problems with fine-grained material at depths  below the
water table were two-fold. Once the air was shut off, water would
rush up the drill pipe to equalize the head between the  formation
and the pipe. As the water traveled up the pipe, it  would often
carry fine sand when it was present into the drill pipe. This resulted
in difficulties with well completion.
  To complete the well, the sand had to be removed from the inner
tube of the drill pipe; this was done either by retracting the drill
pipe until the  sand  fell  out and/or by adding certified clean
municipal water to the drill pipe and forcing the sand out.  The
result was that wells were completed above total drill depths, and
city water had to be introduced into some of the boreholes prior
to completion. However, the desired completion depths never had
to be compromised due to sand in the drill pipe because the desired
depths in flowing sands were achieved  by overdrilling holes and
pulling back.
  The second problem associated with the drilling method in loose
fine sands was  the production of cuttings in excess of the bore-
hole volume. When drilling below the water table in the flowing
or liquid sands, the water produced would carry volumes of  sand
greater than the borehole diameter to the drill bit and out to the
dumpsters.
  Excess fine sand production increased with greater depths below
the water table  as the head difference between the formation and
the drill pipe increased. In one well drilled to  a depth of 234 ft,
with a static water level  at 65  ft,  approximately  20 yd1 of drill
cuttings were produced from a hole that should  have yielded 5
yd3. Approximately 1 to  2 weeks after completing the well,  soils
20 to 40 ft from the well slumped or subsided forming a depres-
sion approximately 5 ft by 40 ft and 2 ft deep.
  Solids were overproduced primarily at the four deep holes drilled
at the site (120 to 250 ft below grade). The produced volume ranged
from 200 to  400% in excess of the drilled borehole volumes. Other
than the surficial subsidence at one well, problems  did not  arise
from overproducing drill cuttings.
  Previous attempts to complete wells in the flowing sands at the
site using augers and mud rotary met with difficulties and either
failed (hole  grouted, reattemptcd, completed at shallower depth)
or were costly  due to the difficult drilling  environment.

WELL DEVELOPMENT
  Monitoring wells were developed by alternately surging and
bailing the wells. This method was effective in developing gravel
360    SITE REMEDIATION

-------
packs in unconsolidated formations while minimizing the volume
of liquids produced. Liquid volumes were of concern due to their
required offsite disposal. Overall, the wells required only minimal
development due to the drilling method used and accurate place-
ment of the gravel pack.

RESULTS
  The dual wall hammer drilling method provided a clean and cost,
effective approach to drilling, sampling and completing monitoring
wells at the San Jose 6 Superfund site. Use of the dual wall hammer
drilling technique resulted in completing  19  monitoring wells,
ranging in depth from 23 to 240 ft, in 20 working days. The origi-
nal project schedule had allowed 50 working days to complete the
wells. Because of the advanced schedule, the drilling subcontrac-
tor was 37% under budget, the overall task was 30% under budget,
and the project schedule was advanced. The total cost for drilling
and well installation,  both CH2M HILL's and the drilling con-
tractor's, was approximately  $257,000 for 1,805 linear feet of
drilling and well completion, or approximately $143/ft. This price
does not include costs associated with off-site disposal of drill
cuttings and  liquids or the analytical costs  associated with waste
disposal.
  In addition to the advanced schedule and resulting lower cost,
the drilling technique succeeded in meeting the following project
objectives  and selection  criteria:

• Cross-contamination between aquifers was either non-existent
  or negligible due to the drill pipe's acting as a drive casing and
  the rapid rates of drilling and well completion
• The drilling method yielded clean holes that required a minimal
  amount of development and  provided representative ground-
  water samples
• Well completions, 4-in., were readily made through the 6-in.
  inside-diameter inner  wall of the drill pipe, thus eliminating
  problems with maintaining bore hole integrity during well com-
  pletion.  There were no difficulties in developing the wells, even
  in fine sand formations, indicating that  the gravel packs were
  properly placed.
• The drilling method provided quality lithologic samples as the
  cuttings were carried from the bit to the cyclone at a rapid rate
  with a minimal amount of mixing. Minor difficulties did occur
  in defining depths within less than 1-ft to lithologic units. This
  was due, in part, to clays temporarily clogging the discharge typi-
  cal of reverse circulation systems and the rapid rate at-which
  the hammer advanced the drill pipe
• Solids and liquids were easily contained in dumpsters. Liquid
  and solid disposal costs using this drilling method were less than
  those associated with mud rotary techniques and were essentially
  comparable to hollow stem auger methods


  Difficulties associated with using the dual wall drilling method
at the San Jose 6 site included production of cuttings in excess of
borehole diameter in fine sand formations below the water table.
This resulted in larger than anticipated volumes of cuttings during
the  drilling of several wells. In one instance, this produced sur-
ficial subsidence adjacent to the well.
  Additional air circulation concerns, specifically health and safety
issues in the breathing zone and the potential to introduce air com-
pressor oils in the borehole, were not major problems. Most of
the work was accomplished using level D health and safety pro-
tection with only occasional upgrades to level C. Water  quality
results did not indicate borehole contamination from oils that could
have been introduced through the air compressor.

CONCLUSION
  As  with any drilling technique, the dual wall hammer  drilling
technique must be applied in the "proper" environment to  achieve
the desired results. Based on our experience at San Jose 6 site and
discussions with Layne Environmental, the following conclusions
have been developed for applying the dual wall hammer  drilling
technique:

• The method is highly effective in drilling coarse sand and gravel
  deposits  above the water table, including  large cobbles
• The method works most effectively in unconsolidated sediments
  below the water table (1) when the sediments are coarse enough
  to prevent them from flowing up the  drill pipe, (2) where the
  formations will stand on  their own, and (3) where the drill depths
  below the watertable are less than 40 ft. In  fine-grained flowing
  sands, wells  can be completed but extended  periods  will be
  required for completion and additional volumes of solids may
  be produced
• By  using a drive casing and completing the well within the drill
  pipe before the pipe is pulled out of the hole, wells can be com-
  pleted  in unconsolidated formations that  sluff or collapse
• The method  also can be used to drill clays,  although these
  materials tend to temporarily and intermittently clog the  dis-
  charge, which  could limit  the  quality of lithologic samples.
  Difficulties with logging clays can be overcome by collecting drive
  samples through the inner wall of the drill pipe
• Since the outside wall of the drill  pipe acts as a drive  casing,
  the method is effective in areas where cross flow between aquifers
  is of concern and/or where lost circulation  would be a problem.
• Due to the direct discharge of liquids and  solids from the drill
  rig through a hose, the method is well suited to sites that  require
  full containment of liquids and solids
• Due to the potential to overproduce cuttings, caution should be
  applied in using the method at depths greater than 40 ft below
  the water table in thick  sequences of fine,  poorly consolidated
  sand formations
• In areas where volatilization of organics or particulates is a health
  and safety concern,  the bins  and/or cyclone could be  vented
  through a stack or large hose away from  the breathing zone

REFERENCES
1.  Kelly, V. C., "Geology of Albuquerque Basin, New Mexico." New
   Mexico Institute of Mining and  Technology, Memoir 33. 1977.
2.  Driscol, F. G. Groundwater and Wells. 2nd Ed. Johnson Division,
   St.  Paul, MN, Johnson Division. 1986.
3.  Barcelona, M. J., J. P. Gibb, and R. A. Miller.  "A Guide to the Selec-
   tion of Materials for Monitoring Well Construction and Groundwater
   Sampling." Champaign,  IL, Department of Energy and  Natural
   Resources. Aug, 1983.
4.  Keely, J. F. and K. Boateng. "Monitoring Well Installation, Purging,
   and Sampling Techniques—Part 1: Conceptualizations." Ground Water.
   25, No. 3 1987.  300-313.
                                                                                                   SITE REMEDIATION     361

-------
                Innovative, Fast-Track  Multi  Media  PCB Cleanup
                                                Kevin  Chisholm,  P.E.
                                                   WAPORA, Inc.
                                                   Rosslyn,  Virginia
ABSTRACT
  A turnkey hazardous waste RI/FS and cleanup was conducted
for a PCB-contaminated impoundment, wetland and stream in a
9-month time frame. The exposure and disposal concerns for the
dioxins and furans present at the site were assessed and found to
be no more significant than those for the PCBs. Several creative
and cost-saving approaches were taken for the investigative and
cleanup phases of the project. These innovative steps included
topographical analyses to minimize RI efforts, development of an
on-site treatment system for wastewater, dewatering of areas with
high water tables, separation of floating oil/sludge from water and
establishing verification of waste quantities for cost accounting.
Various polymers for condensing oil phases were tested and found
minimally useful. In all, over 5,000 gal of oil,  100,000 gal of water
and 2,300 yd3 of sludge and soil were removed and treated or di-
sposed of.

INTRODUCTION
  During the 1960s and early 1970s, a large industrial  facility
deposited its waste PCB-contaminated hydraulic oils and other mis-
cellaneous oils in an impoundment (Figure 1).  From the Impound-
ment, the oils were  either reclaimed off-site or set afire. The
impoundment was constructed with a decanting system, apparently
to drain the water, which separated in a lower phase. However,
occasionally PCB-contaminated sediments, sludge and oil would
be drained with the wastewater. This process resulted in the con-
tamination of extensive downgradient areas including a swale which
conveyed the discharge, a wetland-like area, an intermittent stream
and, to  a low level of contamination, river sediments.
  Due to the past practice of setting the floating oils afire, our
first priority in the RI  was to establish the level of the worst-case
dioxin/furan concentrations at the site. Sample analyses were per-
formed for such a worst-case assessment. All congeners of tetra
through octa  chlorinated dibenzo-p-dioxins  and  dibenzofurans
(CDDs and CDFs) were evaluated. The results showed nondetec-
table levels of the tetra, penta and hexa CDDs and CDFs, with
an equivalent 2,3,7,8-TCDD concentration of  less than 1 ppb. The
only detectable CDDs or CDFs were hepta and octa congeners.
A thorough review and risk assessment with  the available infor-
mation was conducted. Both the state regulatory agency and  the
company being considered for ultimate disposal of the waste were
provided with the findings. All parties agreed that the soil/sludge
was acceptable for landfilling and that the oil was acceptable  for
incineration.
  The following sections describe the highlights of the remedial
investigation,  cleanup plan development and cleanup.

362    SITE  REMEDIATION
                                                                                                      10    20    ao
                            T ....... -
                             I  TfซilmซM Plint
                             I - iF.nc
                                            '
                            V                 >
                             _ Ac cซtซ Road _ ฃ
                                              CO
                                             •••I
                            :     \
                             Tซr    x—Wetland
                                              Landfill
  •    6     ซ     J    t     1     0    10    10   30

                          Figure 1
                          Site Plan
IMPOUNDMENT
  The impoundment's dimensions were 100 ft by 40 ft by an aver-
age depth of 8 ft (Figure 2). The impoundment had an earthen-
raised berm around its perimeter and was  unlined. The waste
material in the impoundment consisted of three phases: (1) floating
oil/sludge, (2) wastewater and (3) sludge and soil. There was no
safe and practical method to determine the depth of contaminated
sludge and soil beneath the impoundment until after the floating
oil/sludge and wastewater were removed.

Floating Oil/Sludge Phase
  The floating oil/sludge was a non-homogeneous mixture of oils
of various viscosities, chunks of paraffin-like compounds and
clumps of decayed leaves which had accumulated over the years.
It was hypothesized that extensive exposure to sunlight and seasonal

-------
                                 too  J
                                            rTanks
                          Figure 2
                        Impoundment

temperature variations, along with forms of bacterial action, had
transformed what had likely been a pumpable floating oil phase
to a mixture of highly viscous oils and floating solids.
  Due to the high PCB concentration (up to  6600 ppm) of the
floating oil/sludge phase, incineration was the required disposal
method. The best methods of removing the floating oil/sludge were
evaluated: (1) direct pumping to a tanker  truck, (2) pumping
through an oil/water separator (or possibly an adjacent above-
ground tank  which was already  contaminated) and (3) miscel-
laneous bulk means of removing  the oil/sludge for placement in
drums. Several factors were considered in selecting a method, such
as: (1) general pumpability; (2) ambient temperature, which affects
pumpability; (3) concentration of solids in the oil/sludge; and (4)
the PCB concentration in the floating oil/sludge versus much lower
PCB concentrations in the sludge at the base of the impoundment.
The cleanup phase was conducted in the summer  months which
made the temperature advantageous for pumping. However, even
under those best-case conditions, the prevalence of solids and high
viscosity made pumping with  a double-diaphragm pump in tan-
dem with a large compressor impossible. The sludge could not be
allowed to drop to the impoundment bottom by simply treating
the water phase since the higher PCB content (< 500 ppm) in the
floating phase would have resulted in the need to incinerate the
bottom solids with their lower PCB concentration  (<500 ppm
PCBs).
  The method of removal finally selected included bulk removal
of floating materials, placement  in a field-rigged separator  for
oil/sludge  and water, and manual loading of the liquid in open-
top drum. To further minimize the volume of material leaving the
site, several combinations of polymers were tested. Unfortunately,
the polymers proved unsuccessful in separating a treatable water
phase. In total, over 4,000 gal of floating oil/sludge were removed
from the impoundment  and incinerated.

Wastewater Phase
  A wastewater treatment system (Figure 3)  for treating and  on-
site discharge of the 100,000-gal wastewater  phase was designed.
The system consisted of a pre filter to remove bulk solids and some
oils, two granular-activated carbon (GAC) units in series, storage
of treated wastewater and discharge via an existing NPDES out-
fall after analytical results showed the presence of less than 1 /ig/1
PCBs. The system's success was predicated on the high selectivity
for the non-polar PCB molecule. The rate of treatment of the
wastewater was determined by an evaluation that considered the
average PCB content of the untreated wastewater, the contact time
needed for that PCB content and the size of the GAC units. Toward
the last stage of the process of pumping wastewater from the im-
poundment, the flow rate was greatly increased to  expedite the
entire  impoundment cleanup phase.  As a result, however, the
wastewater needed to be  treated twice.
4400 gal. link
with — 3000
gals, of oil
o

Empty 4400
gal. tank
Or

Impoundment Approximate
Volume 100.000 gallons
Dimensions •"^100 It. x 37 (t. x S It.
Q - 1 5 gpm

% Inlet Screen
(floating)
	
Q Gage 	
Pump* | J> - 20 pal }

  \
            Klensorb Unit
            1000 Ibs. agent
            2500 Total His.
            H = 67 Inches
            D = 44 Inches
2 x Granular Activated
   Carbon (GAC) Units
2500 Total Ibs.  -each
  1000 Ibs. GAC  ปach
       H = 67  Inches
       O = 44  Inches
                                              ge Basin    /
            200 It. to 150,000 gal. STP Final Olscharg
            (STP Flow = 175 gpm)
            Then to  NPDES outlall 002.
            • IVE*. WITH ASSUMED AVERAGE FLOW OF 4.aซ1O7(p
            Effective dilution Is 3.2 x 10e : 1.
                          Figure 3
           PCB-Contaminated Water Treatment System
Sludge/Soil Phase
  The risks and benefits of investigating the depth of PCB penetra-
tion in the sludge and soil beneath the impoundment were weighed.
The risks were primarily to field scientists and technicians who
would be getting cores of the materials beneath the impoundment
from a boat immersed in the floating oils/sludge. The benefits were
precisely identifying the quantities of contaminated material to be
dealt with. Fortunately, the client concurred that the risks imposed
outweighed the benefits of identifying the precise quantities, since
that would not have  affected the cleanup approach.
  It became apparent after the wastewater phase was removed that
the porous-fluvial types of soils beneath the impoundment allowed
significant penetration and little soil adsorption of PCBs (Figures
4 and 5). Cores were obtained from the bottom and sides of the
impoundment. A diesel fuel component in the waste oils provided
the solvent for carrying PCBs as deep as 6 ft into the sandy soils.
The depth of penetration was also greater where the hydraulic head
was greater. As a result, the zone of contaminated soil was shaped
somewhat like an inverted light bulb. In the removal phase, the
uncontaminated soil at the top of the sides of the impoundment
was removed and stockpiled to keep it from subsiding into the con-
taminated soils.
                                                                                                SITE REMEDIATION    363

-------
 PSSSgs;^^
•  ' •"!  •;• ..;!iV.',:":"v;..'.':;-';.V1.v--- '-'-•'!•.••;•''•'•••  "  ..... '       '
                                                                                                                 XL

                                                                                                                 1*.
                                                                                                WATER TABLE
                       L-..1
                           FLUVIAL DCPO8IIT :  ""• oซ*imo IANO WITH OIICONTIHUOU* CIAT imtis
                                             • CAll
                                              III!
                                                                   Jjtf
                                                           Figure 4
                                              Impoundment Cross-section (side view)
         WATER TABLE
            *• • ••  •^^•fc^iil. L . . •".' ~ '. tW •   •ป!ป••• %>r •> vn. nijiawT nw   • • ;; \, .\fttj




         LjFLUVIAL DEPOSIT :  ""• •DAINIO SANB WITH DKCONTIKUOW* CLAY
                                                tCALI l
                                                 (Ml
                                                                      | 10'
                                                           Figure 5
                                              Impoundment Cross-section (front view)
  The sludge phase was wet enough to require stabilization. The
available dry, clay-like soil on-site was used to stabilize the sludge.
However, approximately 30 tons of lime were ultimately needed
to pass a slump test, thus yielding sludge suitable for landfilling.
Ultimately, over 1000 yd3 yards of sludge and soil were removed
from the impoundment and  disposed of in a TSCA-approved
landfill.

DOWNGRADIENT AREAS
  The impoundment was designed so that the water phase could
be drained off with the use of catch basin and valved drain pipe
(Figure 4). The wastewater was discharged down a swale to a lower
area which had some of the characteristics of a wetland. This lower
area was drained by an intermittent stream which runs about a
quarter of a mile before discharging to a river.
  The remedial investigation effort was concentrated on determin-
ing the outer limits of PCB contamination or the locations where
PCB concentrations decreased to 5 or 10 ppm. It was anticipated
early (Feb. 1987) that the PCB cleanup level would be 25 pppm.
This level was ultimately established in the  U.S.  EPA's  PCB

364    SITE REMEDIATION
                                               cleanup policy for restricted-access areas (Apr. 2, 1987 Federal
                                               Register).
                                                 The presence of a diesel fuel or similar hydrocarbon in the waste
                                               stream acted to carry the PCBs to greater depths in the soil profile.
                                               Based on experience at  previous  PCB-contaminated  sites, the
                                               penetration and migration were much more than would be expected
                                               for PCBs mixed in higher carbon chain fluids than diesel fuel. This
                                               was apparent in both sands and clays.
                                               Swale
                                                 The swale conveyed the wastewaters discharged from the im-
                                               poundment to the lower area. Approximately 150 to 200 yd-1 of
                                               soil were delineated for removal from the swale. The penetration
                                               of PCBs into the soil was as great as 3 ft in some locations.

                                               Lower Area
                                                 The lower area was the largest single area on the site with over
                                               one contiguous area (Figure 6). The lower area had a wetland-like
                                               nature (i.e., elevated water table and vegetation indigenous to wet-
                                               lands). The elevated water table presented difficulty in the cleanup
                                               phase for truck access and sediments too wet to transport. There-

-------
                                                                                                                       z
                                                                                                                       o
                                                                                                                       •n
                                                                                                  Flnil  Depth
                                                                                                  of  Excavation
                                                                                                          
-------
 •  Documentation of cleanup activities (technical and financial—
   possible cost recovery action
 •  Certification of clean site by regulatory agency
   Obviously, several  factors contributed to  such a fast-track
 cleanup for a fairly complex site. The client's representatives recog-
 nized that they had a problem that required resolution, and made
 the  necessary  corporate commitment  without  establishing  an
 adversarial relationship with the regulatory agency. The regula-
 tory agency maintained a reasonable and consistent position. The
 consultant interfaced with the client and  regulatory agency and
 identified the key issues "up front" to minimize surprises, besides
 working diligently to further the project.  Finally, an easily veri-
 fied cost-accounting system was established between the consul-
 tant and the subcontractor to prevent disagreements and delays.

 RESTORATION
   On-site restoration did not commence until after sampling slowed
the site was clean as per the U.S. EPA PCB cleanup policy and
the agency concurred that restoration could begin.
  Restoration basically included grading the site to a smooth topo-
graphy, eliminating  ponding and establishing a hardy, suitable
vegetative cover. To comply with U.S. EPA's cleanup policy, the
site's limited access status is being maintained indefinitely. This
would  preclude nearly all activities in the area.
CONCLUSION
  The entire  RI/FS and  cleanup process  was completed  in
9-months. The entire RI/FS phase was completed for less than
$100,000. This compares with the average RI/FS,  which takes
about 2 years and $650,000 to complete. It is estimated that, by
developing some innovative  approaches to the cleanup process and
maintaining a fast-track time schedule, $500,000 or more was pared
from  the total expenditure.
366    SITE REMEDIATION

-------
                             Design  of Remedial  Measures  and
                                     Waste  Removal  Program
                            Lackawanna  Refuse Superfund  Site

                                                Marcella J. Blasko
                                                Beth F. Cockcroft
                                                William C.  Smith
                                                Patrick F. O'Hara
                                         Paul C.  Rizzo Associates, Inc.
                                            Pittsburgh,  Pennsylvania
ABSTRACT
  The Lackawanna Refuse Superfund Site in Old Forge, Penn-
sylvania includes large areas of uncontrolled land disposal of
drummed liquids and sludges, bulk organic liquids, and consider-
able  quantities  of municipal  waste. This paper discusses the
development of the design and construction bid package for the
remedial/removal program for the site.
  The design began with a value engineering review to identify
special high-cost items in the remedial action design that might be
modified to produce a cost savings. Based on conclusions from
the value engineering review, a supplementary drilling program was
conducted at the site. The project then proceeded through Concept
(30%) and Final (95%) Design with review conferences held after
both of these design phases. The reviews were conducted  by
representatives of government agencies, public interest groups, and
PRPs. Construction cost estimates were prepared at various stages
of the design, with the final estimate based prepared on the adver-
tised Plans and Specifications.

INTRODUCTION
  In June 1987, Paul C. Rizzo Associates, Inc., (Rizzo) completed
a remedial design contract with the U.S. Army Corps of Engineers,
Omaha District (COE) for the cleanup of the Lackawanna Refuse
Superfund Site in Old Forge, Pennsylvania. The remedial measures
were those prescribed in the site ROD issued by the U.S. EPA,
Acting Region HI Administrator.
  Notice for construction of the project appeared in the Commerce
Business Daily  in April 1987. As of July  1987, the remedial
contractor has been chosen, but construction has not yet begun.
The COE has awarded Rizzo a contract to review contractor
submittals and to perform periodic inspections of the construc-
tion throughout the life of the project.

BACKGROUND
  The Lackawanna Refuse Superfund Site is located on a hillside
west of the Borough of Old Forge in Lackawanna County, Penn-
sylvania (Fig. 1). The site, which has an area of 258 acres, has been
strip  mined, and several underlying coal seams have been deep
mined.  Subsequent to the strip  mine  activities,  municipal and
commercial refuse were disposed in open strip cuts on the site. In
1978, the Commonwealth of Pennsylvania Department of Environ-
mental  Resources (PADER) discovered that  on-site activities
included the illegal disposal of industrial and hazardous waste and,
therefore, suspended the solid waste disposal permits.
  Streams and drainage ditches on the site generally drain to the
south and east toward St. John's Creek, which is a tributary of
                   AUSTIN  HEIGHTS
                   SECTION OF
                   OLD FORGE
            PIT 1

         PIT

  LACKAWANNA
  REFUSE  SITE
       PIT  2
       PIT  3
 BOREHOLE PIT
OLD FORGE MINE
POOL  OUTFALL
          OLD  FORGE
VILLA  CORP.
TRAILER PARK
                       Figure 1
                 Lackawanna Refuse Site
                      Plan View
the Lackawanna River. There are no known users of surface water
for drinking water purposes within a downstream distance of 30
miles from the site.

Geology and Hydrogeology
  The site is located in the Lackawanna Valley in the Valley and
Ridge Province of eastern Pennsylvania. The coal found in seams
underlying the valley is referred to as the Northern Anthracite Field
and has been mined since the early 1800s (Fig. 2). The seams have
                                                                                       SITE REMEDIATION    367

-------
been extensively mined by the room and pillar method.
  The natural groundwater regime at the site has been extensively
altered due to deep mining of coal. A consequence of the deep
mining  activities and subsequent  subsidence is  that  extensive
fracturing has occurred in the rock between coal seams. Fracturing
of the bedrock has significantly increased vertical permeability in
what previously had been confining layers, with the result that
precipitation introduced at the surface is able to flow downward
to a mine pool which underlies the center of the valley. Enclosing
the mine pool are barrier pillars (areas of unmined coal left along
company property lines) and collapsed roof rock.
                          Figure 2
                   Generalized Cross Section
                    Lackawanna Refuse Site

  Deeper penetration of the groundwater beneath the mine pool
is prevented by the rock stratum underlying the deepest coal seam.
The mine pool, which discharges into the Lackawanna River,
already is contaminated from off-site mine drainage, and no clearly
site-related contamination has been identified within the mine pool
system.
  The most significant groundwater associated with the site is local-
ized shallow  groundwater flowing in and through the existing
disposal pits. Currently, a portion of this water emanates as seeps
in a few  locations at the base of the landfill areas.

Site Investigation Results
  Investigations subsequent to the one conducted by PADER in
1978, including a U.S. EPA initiated RI in 1983, have indicated
a number of specific problems, the most serious of which is  the
presence of potentially hazardous waste mixed with municipal
refuse in a strip cut designated as Pit 5. Pit 5 reportedly was used
to dispose of approximately 15,000 drums. A substantial portion
of the contents of these drums is believed to be organic solvents.
Additional areas of concern include two other refuse disposal pits
(Pits 2 and 3), an access road to the site on which liquid wastes
were reportedly used for dust control, an area referred to as  the
Borehole Pit where bulk liquid wastes were reportedly dumped and
an  additional small  area where a  hardened paint-like material is
dispersed on the surface. Pits 2 and 3 contain primarily municipal
and commercial refuse, and only trace levels of a few contaminants
have been detected in the access road, Borehole Pit and paint spill
areas. The contamination had penetrated the soil in these areas
less than 1 ft.

Feasibility Study and Record of Decision
   A FS was conducted in 1984 and 1985 for the purpose of defining
potential remedial alternatives for the contamination problem areas
defined during the investigation phase. From the list of remedial
alternatives presented in the feasibility study, the Acting Regional
Administrator of the U.S. EPA Region III selected one remedial
action for each area of concern, as documented in the ROD. The
remedial actions selected are as follows:

• Removal of all drums and highly-contaminated materials from
  Pit 5  for off-site disposal at a qualifying RCRA facility
• Construction of a clay cap over Pits 2, 3 and 5 to meet RCRA
  requirements
• Installation of surface water drainage diversion around all three
  pits and  construction of a leachate collection and treatment
  system for all three pits
• Construction of a gas venting system through the caps of all three
  pits
• Removal of the top layer of contaminated soil from the Bore-
  hole Pit for of f-site disposal at a qualifying RCRA facility and
  returning to grade with a soil cover
• Removal of the top layer of contaminated soil from the access
  road and reconstruction of the road with appropriate drainage
  and sedimentation controls
• Removal of the dried paint and contaminated soil in the paint
  spill area for off-site disposal at a qualifying RCRA facility

VALUE ENGINEERING REVIEW
  In addition to the design effort required to satisfy the ROD,
Rizzo contract also included a value engineering  (VE) review that
was performed immediately after receiving the Notice to Proceed.
The purpose of the VE review was to identify special high-cost items
in the remedial action design that might be modified to produce
a cost savings. The identified items can be such as to allow either
an immediate adoption or be subject to a VE study prior to a final
decision. After a review of the  feasibility study and the selected
remedial action items listed in the ROD, 10 potential value
engineering issues were considered.
  Three issues identified in the VE review had major impacts on
the design team's approach to the remediation. These pertained to:

• Pit 5 subsurface conditions
• Off-site disposal of surface soils
• Cap design

  According to  the ROD, all drums and highly-contaminated
materials are to be removed from Pit 5; and  the selection of highly-
contaminated material to be disposed off site shall be accomplished
by a  large scale sifting through the landfill  waste. The costs were
absent from the FS estimate for the excavation, stockpiling, sifting
and backfilling of uncontaminated  material that surrounds the
highly-contaminated material. The quantity of material in  Pit 5
that  might  have to be handled (Pit 5 was originally more than
100 ft deep) could not be estimated with a high degree of confi-
dence based on the FS information.
  In addition, no comprehensive test boring program and moni-
toring well  installation was ever undertaken in Pit 5, the primary
reason being the perceived risk to personnel and nearby inhabi-
tants. As a  result, the subsurface conditions (i.e., groundwater
conditions, depth of contaminated materials, and types of materials
to be excavated had never been fully defined or documented).
Therefore,  to  better characterize Pit 5, a  VE study was recom-
mended and accepted.
  Soils targeted in the ROD to be excavated from the Borehole
Pit, access road and paint spill area showed minor contamination.
Because the volume of these soils is small (approximately 2,000
yd3) and contamination is minor, it was recommended in the VE
review that U.S. EPA approval should be sought for on-site dis-
posal of this contaminated soil in Pit 5. The U.S. EPA responded
that changing this disposal method would require a re-issue of the
ROD because  one of the basic intents of the remediation, as stat-
ed in the ROD, is the removal of all highly-contaminated materi-
als from Pit 5,  leaving only "clean1 municipal refuse and mine spoil.
Therefore,  no further action was taken on  this VE review recom-
mendation.
368     SITE REMEDIATION

-------
  The ROD states that a clay cap that meets RCRA requirements
shall be constructed over Pits 2, 3 and 5. The 1982 RCRA guide-
lines for final cover design were compared to the cap design recom-
mended in the FS. The cap design was found to be consistent with
1982 RCRA guidelines  with respect to the following:

• Vegetated cover 24 in. thick
• Middle drainage layer 12 in.  thick with hydraulic conductivity
  greater than or equal to 10 ~3 cm/sec plus a filter fabric layer
• Compacted clay layer 24 in.  thick with hydraulic conductivity
  less than or equal to  10~ cm/sec

  However, the  RCRA  guidelines also  specify  a synthetic
membrane between the middle drainage layer and  the clay. The
synthetic membrane is to keep the cap no more permeable than
the liner (RCRA facilities have synthetic membrane liners). Because
no  liner was specified for Pit 5 or Pits 2 and 3,  it was concluded
during the VE review that a synthetic membrane in the cap should
not be required because the clay layer alone fulfills the intent of
the RCRA guidelines and the ROD. Therefore, no formal VE study
was recommended for the cap  design. In summary, it should be
noted that the VE review recommendation for on, site  disposal
of  the contaminated soils would have required a re-issue of the
ROD, but changing from a clay to a synthetic barrier layer would
be  a design variation only,  still fulfilling the intent of the ROD.
PHASES OF DESIGN AND REVIEW
  Rizzo was under contract to the COE, Omaha District from July
1985 to June 1987 to do remedial design work for the cleanup of
the Lackawanna Refuse Superfund Site. The first submittal in
September 1985 was a Preliminary  Construction Cost Estimate.
Concept Design (30%) documents were submitted in October 1985.
These documents consisted of a Design Analysis Report (which
included Outline Specifications), Drawings, a Site-Specific Quality
Management Plan, a Site, Specific Safety Plan (the previous two
documents provide guidance for the COE's review of Contractor's
submittals and performance), a Post-Closure Plan and a Leachate
Assessment Report.  Comments on  the design  documents were
provided by the COE, U.S. EPA, PADER, the Old Forge Toxic
Waste Removal Committee (OFTWRC) and the Old Forge Steering
Committee (representing PRPs). A Concept Review Conference
between the designers and the reviewers was held in December 1985
to  discuss the comments and to decide  on what changes were
necessary to the design documents before final design was initiated.
  Some project work continued through the spring of 1986 (e.g.,
Pit 5 subsurface investigation): however, major Final Design (95%)
effort did not begin until July 1986.  The Final Design documents,
including a Construction Cost Estimate, were submitted in August
and September 1986. Two conferences were held to review the 95%
design, one in Omaha,  Nebraska in November  1986, and one in
Old Forge, Pennsylvania in December 1986.
  Revisions to the design documents in response to the 95% Review
Conference comments were made, and the 100% design package,
exclusive of the Final Construction Cost Estimate, was submitted
in March 1987. The Final Construction Cost Estimate was deve-
loped from the advertised bid package which included all amend-
ments to the Plans and Specifications. The cost estimate submittal
was made in May 1987.
  Most notable throughout the phases of design and review was
the essential  and frequent  communication between  those
responsible for the health and safety aspects of the project and
those responsible for the civil design work. Many in-house project
meetings were scheduled; however, most major design features were
developed at impromptu meetings of the health and safety and civil
design teams. For example, many discussions were held concerning
the daily excavation rate achievable in health and safety equipment
which affected the sizing and layout of the staging areas that will
store materials and drums excavated from Pit  5.
  Not only was effective communication within the design organi-
zation critical to the completion of the remedial design, but effective
communication between Rizzo Associates and the COE and U.S.
EPA also kept the project in focus and on schedule.

SUPPLEMENTAL STUDY
  Based on the VE review, the COE decided that a subsurface
investigation of Pit 5 should be performed to obtain information
needed for a safe and cost-effective design of the remedial/removal
program and to enable a more accurate determination of the cost
of remediation. Because of the legitimate concerns for public safety,
a specialized  drilling technique was designed using technology
adopted from the oil and gas industry originally developed for well
installation through formations containing naturally occurring pres-
surized toxic gases. A site-specific program for personnel protection
also was developed and implemented.
  The  supplemental program  was designed in January  1986
between the Concept and Final Design phases and drilling was con-
ducted in February and March 1986. For  a detailed description
of the Lackawanna Pit 5 drilling program, see the paper by O'Hara,
atoll.1
  The following information resulted from this investigation and
was used in completing the design process:

• The  depth  of refuse disposal in Pit 5  is relatively  uniform,
  varying between  15 and 19.5 ft below ground surface. This in-
  formation was used to estimate excavation volumes which affect
  backfill requirements, staging areas,  analytical tests, etc. Pre-
  vious estimates of disposal depths ranged up to  50 ft.
• Groundwater in the pit is probably in perched zones, based on
  water levels and water quality. The volume  of water is impor-
  tant  for excavation dewatering,  and  the quality of water was
needed for disposal cost estimates  and  treatment system design.
  There was no previous detailed information on water levels and
  water quality within Pit  5.
• Mine spoil underlying the refuse is not highly contaminated, and
  significant contamination  generally is confined to the refuse zone.
• Overall  analytical results are indicative of "typical" refuse.
  Pit 5 refuse was not found to vary significantly from Pits 2 and
  3 refuse in  content and concentration.
• A pocket of groundwater within the pit showed the influence of
  possible drummed materials. Two groundwater samples con-
  tained contaminants at levels greater than would be expected in
  a sanitary landfill. The other three groundwater samples were
  considered to be consistent with what could be expected within
  a sanitary landfill.

COMMENT AND RESPONSE PHASES
  As stated previously, a Concept (30%) Review Conference and
two Final (95%) Review Conferences were held to evaluate the
design of remedial measures at the Lackawanna Refuse Superfund
Site. At both Concept Review and Final Review, 18  separate
reviewers representing the COE, U.S. EPA, PADER, OFTWRC
and the PRPs submitted written design comments that were dis-
cussed at the  reviews.
  Specific issues that generated the majority of discussions at the
Review Conferences included:

• Mine subsidence

• Temporary versus permanent designs
• "How clean is clean"
• Disposal options for highly-contaminated materials
• Cap design

  Salient items of the discussions and how the issues were resolved
are presented in the following  paragraphs.
  A major design consideration is the potential of future subsi-
dence due to past deep mining activities beneath the site. During
the remedial  investigation, some borings had  penetrated  the
underlying coal seams and found  some in-place coal and some
voids. One reviewer wanted an extensive drilling program to map
                                                                                               SITE REMEDIATION     369

-------
the mined-out seams and develop remedial measures to prevent
future subsidence. Other reviewers were skeptical of the cost versus
benefits that would be derived from such a program. A series of
deep mining maps showed that  the pillars of all the coal seams
beneath the pits could be removed. In addition, according to the
Office of Surface Mining in Wilkes-Barre, probably only a minor
number of coal stumps  were left  in place for safety during the
extraction operation. Therefore, the design team concluded that
future subsidence, if any, will be tolerable; and with enhanced post-
closure inspection and maintenance the functional behavior of the
cap system will not be impaired. The COE determined that fur-
ther investigation of subsidence potential  was not justified from
a cost/benefit viewpoint.
  Temporary structures and utility service lines will be constructed
and installed at the site for use during Pit  5 excavation and disposal
activities only. Many of the reviewers felt that this construction
should meet  all applicable codes. The design team and  other
reviewers felt that since these structures are temporary, the code
requirements could be relaxed  and acceptable performance still
could be maintained. For instance, a temporary, non-potable water
line was designed initially to supply water to the decontamination
areas and staging areas.  Some reviewers insisted on 5 ft of cover
over the water line (versus the 38 in. initially specified), fire hydrants
on the system, pressure and flow tests on the hydrants,  chlorine
residual tests, etc. The  majority of these comments were later
incorporated in the design.
   There was much discussion at the review conferences concerning
the "how clean is clean" issue. The ROD states that Pit 5 excava-
tion shall include off-site removal of all  the drums and all highly-
contaminated  refuse. Refuse determined to  be equivalent  to
"normal" municipal refuse would be backfilled and contained in
the pit. Therefore, backfilled material in Pit 5 would be equiva-
lent  to a normal sanitary landfill. Based on the results of the
Pit 5 investigations,  Rizzo proposed methodology employing key
indicator compounds (specific  volatile organics)  as  a way  to
differentiate between the highly-contaminated municipal refuse
requiring removal and the normal municipal refuse which will be
contained on site. This key indicator compound (KIC) approach
was outlined in the ROD, and its appropriateness confirmed during
the Pit 5 investigation performed by the design team. The reason
for this approach is the impracticality of using a full Hazardous
Substance List (HSL) analysis as a field decision-making tool. The
environmental, schedule and cost problems associated with long-
term storage of large quantities of materials while awaiting full
HSL laboratory results have contributed to the acceptance of the
KIC approach.
  Some commentors disagreed with the use of KICs and developed
a series of written comments during the Final Design phase con-
cerning the fundamental KIC approach. After significant dialogue,
many of the comments were retracted after the Final Design
Review.
  Another salient issue resulting from the Final Review Confer-
ences was the  suggestion to permit a variety of disposal options
for materials requiring off-site disposal. The ROD had specified
incineration  for all  Pit  5  material unless there was inadequate
incinerator capacity to accept material from the site. At  the time
of the construction solicitation, U.S. EPA Region III determined
that there was not sufficient incineration capacity to accept the
expected quantities and types of materials anticipated. Subse-
quently, the specifications for off-site disposal were modified to
permit a variety of disposal options including incineration, solidifi-
cation, treatment and landfilling.
  Clay caps over Pits 2, 3 and 5 were specified in the ROD. As
previously discussed, the VE review concluded the clay alone (i.e.,
no synthetic membrane) was acceptable for the cap. Some reviewers
agreed and others disagreed, wanting a clay/synthetic composite
cap.  However, the principles for a single  barrier layer  were
sustained. In addition, only one clay supplier was found in the
project area. Thus, it was suggested that if clay were scarce, a
synthetic membrane should be specified as an alternate to the clay.
The U.S.  EPA and COE concluded that an alternate synthetic
membrane cap should  be  included in the design.
  Overall, the comment and response phases were time-consuming,
but the Final Design and bid package reflected the benefits of the
input  received from the many interested parties.


FINAL CONSTRUCTION COST ESTIMATE
  There were eight bids submitted  on  the  Lackawanna Refuse
Superfund Site cleanup. The average of the bids was $25 million;
the median was  $26 million. The design team's estimate of $27
million was within 10% of four of the bids. The low bid was sig-
nificantly less than the estimates obtained during  design.

CONCLUSIONS
  The following conclusions can be drawn from Rizzo's experiences
on the design of the Lackawanna Refuse Superfund Site remedial/
removal program:

• Those involved with a Superfund project should be aware that
changes,
  which are engineering design variations, may be made to the
  ROD without  requiring reissue of the ROD
• The A-E should be aware that even though structures and utility
  services required for the cleanup may be temporary, they may
  have to be designed to  the standards  required  for  permanent
  structures
• The A-E should be aware that the comment and review phases
  may be time, consuming, but their final product will be improved
  overall
• Most importantly, for a Superfund cleanup  design project to
  run smoothly, frequent communication between the A-E, the
  COE and the  U.S.  EPA is essential

ACKNOWLEDGMENTS
  The authors appreciate the support and encouragement from the
U.S. Army Corps of  Engineers, the U.S. EPA, Region III, the
Pennsylvania Department of Environmental Resources, the PRPs,
and the citizens of Old Forge, Pennsylvania in the design of the
Lackawanna Refuse Superfund Site cleanup.

REFERENCE
1. O'Hara, P.P., Bird, K.J.  and Baugham, W.A., "Proceedings of Uit
Seventh National Conference on Uncontrolled Hazardous Waste Sites",
Washington,  DC,  Nov.  1986. 126-131.
370     SITE REMEpIATION

-------
                           Horizontal  Radials for Geophysics  and
                                  Hazardous  Waste  Remediation

                                               Wade  Dickinson, Partner
                                            R. Wayne Dickinson, Partner
                                                   Petrolphysics, Ltd
                                               San Francisco, California
                                         Peter A. Mote, Principal Geologist
                                            Jerome S. Nelson, Consultant
                                             Harding Lawson Associates
                                                   Novato, California
ABSTRACT
  Radial wells can now be placed horizontally or at shallow angles
from a central vertical well to intercept, to inject or to monitor
fluids and gases. This new technology has  great potential in a
diversity of applications: hydrocarbon production; contaminated
groundwater, toxic gas and mineral leachate recovery; geophysi-
cal diagnostics qf hazardous waste sites and domestic water supply
development. Using this technology, a 1-1/4-in. steel drill pipe is
pulled hydrodynamically around a 9- to 12-in. radius turn and ad-
vanced up to 200+ feet laterally to develop the radials in the earth.
In effect, the system drills and cases as  it moves forward.
  Multiple radial tubes can be placed at 5 to 10 degrees spacing
in azimuth at the same level and from one or more levels so that
significant access to the subsurface can be gained from a single
vertical well. These multiple radial capabilities can greatly enhance
efforts to detect, monitor, extract or contain subsurface  con-
taminants, particularly in  populated areas. This paper describes
the methods by which horizontal radial wells are installed. It also
discusses potential  radial  applications to  remediating  and
monitoring hazardous waste sites.

INTRODUCTION
  Conventional methods of control groundwater and subsurface
contamination typically include using vertical wells. However, ver-
tical wells have inherent limitations and difficulties in many geo-
logic formations. Depending on  the  surface and  subsurface
conditions or accessibility of target zones,  factors such  as fractured
and complex geology, hydrogeology, geologic structure, aquifer
permeability and cross-contamination of aquifers can impose
serious limitations on the efficiency of vertical well systems. Radial
wells, on the other hand, are more flexible with respect to accessing
specific subsurface conditions and  structures.  They can provide
viable and cost-effective solutions to the limitations of vertical well
systems.
  The technology of placing horizontal or lateral radial wells has
been under development in the oil industry for  decades. Deviated
drilling from the vertical to the horizontal has been available using
large (1800 to 3000 ft) radius and medium (20  to 40 ft) radius of
curvature to attain the desired horizontal orientation. But such
placement has been limited because only one radial can be placed
in each vertical well, and precise vertical placement of the horizontal
wells has not been possible with such large radii of  curvature.
  The potential advantages of precise radial placement at specific
depths  stimulated Petrolyphysics and the Bechtel Group, both of
San Francisco, to develop and commercialize a novel high-pressure,
hydraulic jet drilling system. The Petrolphysics Radial Placement
System allows the precise placement of horizontal radial wells at
a 9- 12-in. radius of curvature in vertical wells of 4-1/2 in. or larger
casing diameter. In that associated research and development work,
over 500 radials totaling more than 27,000 ft have been placed at
depths from a few feet to 6800 ft.
  Although the  12-in.  radius  of curvature technology  was
developed for the oil industry,  it has potential application  in
hazardous waste projects as well. For instance, placement of many
horizonal radials into a contaminated aquifer from a single con-
ventional vertical well can markedly enhance  access  to the
formation. Placing wells in the plane of an aquifer or perpendicular
to fracture  zones can significantly increase the area of exposure
to the formation as compared to that which can be attained using
vertical wells. With the intrinsic geometric advantage that radials
provide, the volume of formation that can be serviced by a single
vertical well is increased. Hence,  fewer vertical wells are  needed,
and a more uniform and thorough accessibility to the contamina-
tion is available.
  This paper describes the short radius horizontal radial placement
and completion systems and their application to hazardous waste
monitoring, control and remediation.

HORIZONTAL RADIAL DRILLING SYSTEM
  Over the past 7 years, Petrolphysics Ltd. and Bechtel Group have
been developing the technology to place and complete horizonal
radials  at both shallow and great depths. The technology  was
developed primarily to enhance oil recovery from shallow and deep
oil reservoirs. Horizontal wells (radials) may be placed in a pattern
at different  depths in a single well. These radial pipes extend out-
ward from a central well for up to 200 + ft. To place an array  of
shallow radials, as shown in Figure 1, a whipstock is lowered into
a vertical cased bore hole of about 30 to 36 in. diameter. The whip-
stock is attached to a 2-7/8-in. or 3-1/2- in. conventional working
string. A sketch of the shallow Horizontal Drilling System arrange-
ment is shown in Figure 2.
  A 100 to  200-ft long section of 1-1/4-in. diameter  continuous
pipe (Electric Resistance Welded (ERW) tube) is placed within the
2-7/8-in. or 3-1/2-in. working string. That  ERW tube has a
hydraulic drill head welded to its leading end (nose). Once the whip-
stock is placed, drilling is begun by pumping high pressure water
down the working string and out the 1-1/4-in. ERW drill pipe. The
resultant hydrodynamic force on the drill head pulls the 1-1/4-in.
ERW drill pipe through and around the whipstock at a 12-in. radius
so that it enters the formation horizontally (Fig. 2). The high pres-
sure water is accelerated to high velocity by the drill head so that
it bores a hole of selected size (4 to 24-in. diameter in unconsoli-
                                                                                              SITE REMEDIATION     371

-------
                           f
                           T
                                   VERTICAL  WELL
                                           RADIAL
            FOUR RADIAL  AERIAL  VIEW
           TWELVE RADIAL AERIAL VIEW

                           Figure I
                  Typical Aerial Radial Patterns
                 VERTICAL WORKING STRING
                         ^•/.•SEALED CATO*lEWr

                         ^'vV;V::;Vv':::   DEPTH OF RADIAL ARRAY
                                          10 :'.::?;'.':••.'?'.'•:•••:
                          ^lili^- FLUID SUMP
                               ';/-.•'"

                                 CEMENT PLUG
                            Figure 2
                   Shallow Horizontal Drilling
dated soil) into which the 1-1/4- in. ERW drill pipe may move.
As many as 100 radials have been placed in a single well at three
different levels using this technique.

DRILLING SYSTEM  DESIGN
Conical Jet Drilling
   A major part of the $25,000,000 research, development and com-
mercialization drilling program of the Horizontal Radial Project
has involved development of a new concept in hard rock water jet
drilling without abrasives, called Conical Jet Drilling. These nozzle
systems produce a conical shell of high velocity water. This coni-
cal shell of fluid creates  a body of cuttings slurry which appears
to act as an in situ abrasive slurry. Both hard and crystalline rocks
(granite and basalt) and unconsolidated sedimentary rocks are cut
rapidly. Bore hole diameter and direction of drilling may be con-
trolled.
   The Conical Jet drillhead penetrates the following typical materi-
als in the laboratory at atmospheric backpressure as indicated in
Table  I.

                            Tible 1
        Penetration Cipibllllles of the Conic*! Jet Drillbead
                   40000 ftl
      hull          (t.t.)
      (I Ultimo. WA,

      CoorM         17)00 ป•!

      |rult* oa4 tMTtz
      (CuadUii Skl.U)
                                                                       3. loon|*ii*oaa      16)00

                                                                         Crซnitซ
    4. liucl* QMlk    MOO ป000
      (Ctrollto*. TI)
                                                                       3. Stovona S*nซ'stoao
                                                                         (Elk SU1ป. CJ)
                                                                                                  1/1 foot ftr ซlปซti.   aOOO ffl. 134 am
                                                                                                                   MOO ป•!. 134 (•>
                                                                                                  1/2 toot pปr
                                                                                                  U acป> of 1
                                                                                                  loot r
                                                                         (tbn Dh.bl)
                 MOO >•!. 15* I


I* ucซm> of 1  foot    ง000 fml. 134 |


I* OCM. of 1  foot    MOO ftl, 134 |
Drilling and Drill Siring Propulsion
  Drilling of the earth formation and propulsion of the l-l/4-in.
ERW drill pipe use the same hydrodynamk force. That force con-
currently pulls and pushes the l-l/4-in. ERW drill pipe forward
while it cuts the horizontal radial bore hole in the earth. In effect,
the drill string makes its own hole and jumps into it, hence it is
called the "Rabbit Force." It also concurrently cases that bore hole.
  Typical velocities for the drill pipe range from 5  to 120 ft/min
in unconsolidated formations.

Electronic Positional Logging
A highly  flexible wireline positional logging system has been
developed and applied to measure and print out the location of
the horizontal radials.
  The most important features  are: (1) the flexibility (90 degree
turn within  the l-l/4-in. ERW drill pipe at 12-in. radius), (2) the
basic mechanisms of measuring curvature, and (3) the pump down
features. The tools can be used in tubing diameter from 1 in. and
larger and at pressures of many thousands of lb/in.2

Completion Technology
  To provide sand control in horizonal radials or deviated wells,
Petrolphysics has demonstrated the following applications in the
laboratory  on 80-ft  full-scale  radials:  (1)  bidirectional  gravel
packing to 100% fill, (b) electrolytic cut off of the radial down-
hole, (c) electrolytic perforation in situ, (d) placement of flexible
slotted liners, and (e) provision of sand filters on the cut-off ends
of the radials.

GEOPHYSICAL INVESTIGATIONS
Physical Basis
  Site characterization for remedial and hazardous waste inves-
372    SITE REMEDIATION

-------
tigations frequently has been augmented by the use of geophysi-
cal methods. The scientific basis for these methods is that geologic
and contaminant conditions of the soils often directly affect their
seismic velocities (P and S waves), acoustic attenuation, electrical
resistivity, conductivity, temperature and radioactivity. These ef-
fects are the result of changes due to weathering, faulting, frac-
turing,  hardness,  porosity, permeability, saturation, moisture
content and groundwater quality.

Previous Approaches
  Previously, geophysical measurements involved obtaining data
from the ground surface or between the ground surface and loca-
tions at depth in vertical drill holes or between two or more ver-
tical drill holes.  Each situation requires  that  the energy be
transmitted  from  one location  and  be  received by sensors or
detectors at a remote location. However, the geophysical cover-
age of a site with any sensor is subject to the limitations (i.e., sub-
surface measurements and vertical holes) that are imposed on the
usual geologic and hydrogeologic methods of investigations. Thus
a gap in the complete investigation of the site exists along a plane
at a specified depth beneath the site and between that plane and
the surface.

New Concepts for Horizontal and Geophysical Arrays
   With the  advent of accurately positioned horizontal or radial
bore holes, the horizontal information gaps mentioned above can
be closed. Placement of transmitters and receivers  in horizontal
bores will allow measurements to be obtained throughout the sub-
surface space that essentially surrounds the  site and target of
interest. This means that data previously obtained  either on the
surface or between vertical holes can now be augmented by data
obtained between  the surface and some  depth of interest via the
total length of one or more horizontal bore  holes. Thus, a three-
dimensional study can be made on the total block of soil and rock
defined by the limits of the site.

Instrumentation
  Transmitters and receivers of 1/2 to 1 in. in diameter or smaller
have been produced by the geophysical instrument manufacturing
industry. Appropriate circuitry and cabling can be made for their
insertion into the horizontal holes. These transmitters/receivers can
be designed with specific frequency or range of frequencies suited
for specific target resolution. Thus, appropriate geophysical data
can be obtained from within the 1- 1/4-in. horizontal radial tubes.

Conceptual Examples
  Three possible examples for the combined use of horizontal
radial bore holes and geophysical arrays  are shown  in Figures 3,
4 and 5.
  The first example, shown in Figure 3, is for characterization of
a proposed site to consider the use of seismic measurements. This
application would make use of a deeper bore hole and deeper
radials than shown in the Shallow Horizontal Drilling concept in
Figure 2. Deeper applications would allow the oil industry to place
100- to 200 ft horizontal radials at up to 6800 ft depths. With such
deeper horizontal radials, the site could be explored with a mini-
mum of drill  holes by placing a vertical hole directly in the center
of the site with an array of radials radiating out at a selected depth.
Transmitters  and/or receivers would be placed in these radials.
Depending on the number of radials, a series of cross-sections can
be developed  along most any alignment across the site. Thus, sub-
surface conditions between the radials and the surface relating the
rock, fracturing and weathering can be determined for any number
of vertical cross-sections. The seismic travel time between trans-
mitter and receiver (ray path) can be further treated using appro-
priate sectioning  (tomographic) algorithms that  will  identify
changes in velocity occurring along each increment of the ray path
and result in very detailed resolution of subsurface conditions. In
effect, the use of radials  will permit much  better near-surface
seismic ray tracing and data interpretation.
                           Figure 4
           Monitoring of Leachate Under Existing Landfill
                   With Radial Resistivity Array
Existing Landfill
  The second example, shown in sketch Figure 4, depicts a con-
cept for laying out electrical sensors along a series of shallow
radials. Here the radials would be placed by the Shallow Horizon-
tal Drilling concept as shown in Figure 2. Such radials would be
placed either within the landfill material or at a selected depth be-
low it. Material within the landfill could be characterized by the
measurements of its electrical resistivity or conductivity. Data on
the water table, perched water zones, permeability and water quality
could thus be obtained. In arrays set below an existing landfill,
the possible leakage or seepage of fluids from the landfill could
be detected.
                                                                         SURFACE INSTRUMENTATION'
                                                                                                     VERTICAL BORE HOLE
                            Figure 3
            Seismic Raypaths with Radial Receiver Array
                                                                                                SENSORS  IN RADIALS
                           Figure 5
           Cross-Section of Resistivity Monitoring Array
                                                                                                   SITE REMEDIATION     373

-------
Monitoring Array
  Figure 5 shows the use of sensors placed along shallow radials,
again using the Shallow Horizonal Radial concept shown in Figure
2. Here the radials could be placed beneath a new pond to moni-
tor  the performance of a  liner system. The radials with sensors
could be placed in a grid manner to locate where leaks and seeps
occur.  Through  the use of on-site data processing, the informa-
tion also can be channeled into an alarm system for early warning
of any changes in subsurface conditions.
  This concept also can be adapted to a leach pad operation for
mineral recovery. For this purpose, sensors in a radial pipe beneath
the leach pad can be used to monitor the cyanide fluids migrating
through the leach pad. It also is possible to place these radials with
sensors downgradient  from the  facility to monitor any new or
existing conditions related to contaminants seeping or leaking from
that structure. This approach also can be used to  define changes
such as the size or depth of a plume and its direction of migration.
  These three concepts shown in Figures 3, 4 and  5 are but a few
of the potential applications of the use of radials for geophysical
work.  It seems likely that the experience of others with specific
problems will lead to other conceptual approaches to application
of  radials to help define geological problems, to determine site
characteristics and to monitor the changes that occur for each task.

APPLICATION TO  HAZARDOUS WASTE CONTROL
AND REMEDIATION
  The optional applications of controlling migration of hazardous
subsurface waste using the radial well system are numerous. The
key is that an array of horizontal radials provides accessibility to
the subsurface not available with vertical well systems. Some areas
where  application could be very beneficial include:

• Contaminated groundwater recovery and control
• Biodegradation of organic contaminants
• In situ stripping of volatile contaminants
• Injection/extraction from  dense population  centers or from
  beneath civil structures
• Injection/extraction in geologic conditions that are difficult to
  access with vertical systems (e.g., fracture permeability)
• Injection/extraction from aquifers of less than 200-ft thickness

The more  obvious potential applications of horizontal  radial
systems are in the areas of groundwater control and in situ modifi-
cation  of the ground or hydraulic conditions. The benefits of the
system arise primarily from the ability to provide  access to ground-
water, geologic structures and physical situations that cannot be
efficiently provided by vertical bore holes.

Groundwater Control
  To exemplify  the potential advantages of horizontal radials
systems over vertical well systems, some general comparison can
be made. The typical well-field extraction scheme incorporates an
array of vertical wells that intercepts and withdraws the contami-
nated groundwater for treatment  (Fig. 6).  The design of the well
field is dependent on site- specific characteristics such as aquifer
transmissivity, plume geometry,  groundwater and contaminant
characteristics.  Surface access limitations  include steep topogra-
phy, water bodies, wet lands, industrial facilities,  areas restricted
by security, urban developments, streets, highways and powerline
corridors.
  A vertical well  draws water from all directions within the cone
of depression (Fig. 6). To ensure that the entire contaminated zone
is influenced by a vertical well field, the cones of depression must
necessarily overlap into areas of clean groundwater. Thus exces-
sively large volumes of water may be generated for treatment.
  In contrast, the horizontal radial well configuration (Fig. 7) can
be designed so that the exposure of the well  screens  is concentrated
in the contaminated areas. A central vertical well can be strate-
gically placed to deploy radial wells within a contaminant  plume
and/or to avoid surface obstacles mentioned above while attaining

374    SITE REMEDIATION
 WATER
                                          >jSi VERTICAL WELL
                            Figure 6
                 Conventional Vertical Well Field
                                                       RADIAL
                                               CAL WELL
                           Figure 7
             Well Field With Vertical Well and Radials

efficient aquifer coverage. Extracting fluids from aquifers using
the horizontal system can provide economies in capital, operating
and maintenance costs.
  Vertical well fields often require a large number of closely spaced
wells, the installation, capital and operating costs of which are
typically high. Such well fields also require long operational periods
because extraction is often very inefficient.
  The alternative to a field of many vertical  wells  is to use hori-
zontal radials from fewer vertical wells. This approach can pro-
vide significant cost savings because exposure to the subsurface
with radials is more efficient and concentrated within known con-
taminated areas. Reducing the number of vertical wells propor-
tionately also reduces the number  of pumps, the piping and the
maintenance  required.

Subsurface Drains—Lcachate Interception
  A horizontal radial system can  control seepage  from beneath
existing surface and shallow waste facilities (Fig. 8). The horizontal
radials can be emplaced to intercept leachate  from  these facilities
before it migrates away from the site or mixes with groundwater.
  The radial  well system also can be used to monitor the unsatu-
rated zone beneath a disposal site. The configurations that radials
can provide are beneficial to the installation of leachate or vapor
detection devices.  Such monitoring systems  can be designed to
monitor isolated sections within the radials so that one could
determine contaminant concentration gradients across a site.
  In disposal areas where migration pathways are primarily along
fractures (Fig. 9), perimeter horizonal radial arrays  can efficiently
intercept fracture systems to create a hydraulic sink for migrating

-------
                                     EXISTING DISPOSAL TRENCH
                            Figure 8
                 Leachate Monitoring or Collection
                      By Horizontal Radials

  VERTICAL FRACTURES :71^"f} ;*-/i
                            Figure 9
            Horizontal Radial System in Fractured Media
              Vertical Well System in Fractured Media

contaminants. The effectiveness of intercepting vertical fractures
with horizontal radials for improved oil production already has
been successfully demonstrated commercially.
  Other optional uses of radial wells installed as subsurface drains
include: (1) developing  a  positive  pressure gradient towards the
storage site to inhibit contaminant migration (this could be done
using either gas or water);  (2) grouting to seal migration pathways;
and (3) freezing the formation to  seal fractures.

In Situ Treatment Using  Radial Wells
  Horizontal radial systems have the unique ability to penetrate
contaminated soil and ground water in lateral planes, regardless of
depth below ground surface. Therefore, they can be used for a wide
variety of in situ  treatment methods. Some of these methods
include: (1) injection to neutralize alkaline or acidic contamina-
tion; (2) injection of bio-organisms and, perhaps more importantly,
introduction of nutrients for the organisms, and (3) physical modi-
fication by freezing  or grouting.  The latter two methods are
promising as contaminant control methods.
  Bioremediation is a developing method for in situ treatment of
certain petroleum and volatile organic compounds. Because con-
taminants commonly migrate along preferred pathways within
geologic formations,  radials can very efficiently introduce  bio-
organisms and nutrients within the contaminated zone. The regular
supply of nutrients to the organisms is important to the success
of bioremediation. Radials can be maintained for this purpose.
  In situ treatment by solidification/stabilization can be done by
binding the contamination into a solid mass. Solidification can be
accomplished by injecting grout (cement or polymer) or by freezing
the formation by continuously circulating refrigerant or injecting
cryogenic liquid. The radial well system can effectively be used to
solidify a formation.
  Grout can be injected under pressure through the system that
has penetrated the formation to seal potential pathways.  In situ
freezing can be accomplished by injecting cryogenic fluids (nitro-
gen, liquid nitrogen or liquid carbon dioxide).
  The application of radials to control the effects  of subsurface
emergencies at major industrial plants could be of great benefit.
Generally, at  such facilities, the foundation geology is well
documented as the result of permitting and licensing requirements.
The radial well system, which can be placed remotely and rapidly,
could be designed to stop or inhibit the spread of contaminants
away from the accident site.

CONCLUSIONS
  Application of horizontal radial systems offers another dimen-
sion in the subsurface control and monitoring of contaminants in
both unsaturated and saturated materials. Many radials can be
placed from a single  vertical well. They can  be placed in one
horizontal or lateral plane or in several stacked layers. The total
radial array installation  can be done very rapidly.
  The radial system can be emplaced around the periphery and
beneath sites of contamination to provide for  boundary control
and monitoring. Immobilization by in  situ grouting or freezing
using cryogenic fluids could be done efficiently and effectively using
the radial system. In emergency situations, radials will be rapidly
installed beneath a failed and leaking plant to  control migration
pathways and potential leakage off the site.

ACKNOWLEDGEMENT
  The authors gratefully thank the Bechtel Group  Inc., Bechtel
Investments Inc. and  all of their  colleagues  at Petrolphysics,
Harding Lawson and the Mechanical Engineering Department at
University  of California at  Berkeley  for  their support  and
encouragement of this work.

REFERENCES
1. Pendleton, L. E. and Ramesh, A. B., "An Innovative Method of Drilling
  Horizontal Boreholes," Heavy Oil and Oil Sands Technical Symposium,
  Calgary, Alberta, Feb. 1985.
2. Dickinson,  W. and Dickinson, R. W., "Horizonal Radial Drilling
  System," SPE 13949 Proceedings Society of Petroleum Engineers (SPE)
  1985 California Regional Meeting, Bakersfield, CA, 1985.
3. Dickinson, W., Anderson, R. R. and Dickinson, R. W., "A Second-
  Generation Horizontal Drilling System," IADC/SPE 14804 Proceedings
  of   1986  IADC/SPE  Drilling   Conference,  Dallas,   TX,
  Feb. 1986.
4. Simmons, R. N., "Turning the Corner," Oil and Gas Investor, June,
  1986.
5. Crosby, T. W., Head, H. N., Dickinson, W. and Dickinson, R. W.,
  "Management of Uncontrolled Hazardous Waste Sites," SUPERFUND
  '86, Vancouver, B.C., Canada, Dec. 1986.
6. Dickinson, W., Dickinson, R. W. and Mote, P. A., "Radial Wells and
  Hazardous  Waste Sites," Proc. Hazardous Waste and Hazardous
  Materials, Washington, D.C., Mar.  1987.
7. Dickinson, W., Wilkes, R. A. and Dickinson, R. W., "Conical Water
  Jet Drilling." Fourth U.S. Water Jet Conference, Berkeley, CA, Aug
  1987.
8. Dickinson, W., Anderson, R. R. and Dickinson, R. W., "Gravel Packing
  of Horizonal Wells," SPE 16931. SPE Annual Meeting, Dallas, TX
  Sept.  1987.
                                                                                                   SITE REMEDIATION     375

-------
                                A  Model  for  Estimating  the  Cost
                                 Of  Superfund  Remedial  Actions

                                                   Richard K. Biggs
                                      U.S. Environmental Protection  Agency
                                          Hazardous Site Control  Division
                                                  Washington, D.C.
                                               R.  Benson Fergus, P.E.
                                                     CH2M HILL
                                                   Reston, Virginia
ABSTRACT
  A model to estimate costs of remedial actions has been developed
to obtain outyear remedial action cost estimates for Superfund sites.
The cost of remedial actions (CORA) model is microcomputer-
based and has two components: an expert system and a cost system.
The expert system interacts with the user to develop a range of
reasonable  response actions. The cost  system contains  cost
algorithms capable of developing order-of-magnitude cost estimates
for 34 technologies.
  The CORA model was used recently to develop cost estimates
for 97 Superfund sites considered to be possible FY  1989 remedial
action candidates. The U.S. EPA used these cost estimates in the
development of the FY 1989 Superfund remedial action budget.

INTRODUCTION
  Capital and operation and maintenance (O&M)  cost estimates
are required for Superfund remedial actions at sites included in
the U.S. EPA's Superfund Comprehensive Accomplishments Plan
(SCAP). These estimates are used to manage current year activi-
ties and to develop outyear budgets. SCAP estimates also form
the basis for analytical studies that evaluate the impacts of policy
changes. In the past, SCAP estimates were derived from a variety
of sources ranging  from detailed construction bids to estimates
based on limited information about site conditions. Budgets are
developed 18 months prior to the SCAP operating year. There-
fore, the U.S. EPA needed a method to estimate remedial action
costs in the pre-feasibility stage  of analysis. This method was to
incorporate: (1) a reproducible and consistent method of applying
the remedy selection guidance and (2) a straightforward method
to develop  site-specific  order-of-magnitude cost estimates. The
CORA system was developed in response to this need.
  The Cost Remediation Actions (CORA) system is structured to
arrive at a range of representative response actions for a Super-
fund site and to develop preliminary order-of-magnitude estimates
of capital and O&M costs for outyear budgeting. The U.S.  EPA
intends to make the CORA system available to each  regional office
to help standardize the process  for developing outyear remedial
action  budget estimates.

SUMMARY DESCRIPTION OF THE CORA  MODEL
  The CORA model includes two distinct subsystems: the expert
system and (2) the cost system. Both subsystems operate  on a
microcomputer and are independent of each other.
  The expert system asks questions and functions as a knowledge-
able advisor. Based on the user's answer to each question, the sys-
tem determines what additional facts are required to arrive at a
recommendation for the site. It pursues viable lines of reasoning
based on technical feasibility and the U.S. EPA policy considera-
tions, which lead to site cleanup recommendations.  The expert
system contains questions for three major areas of contamination:
(1) soils, sediment and mixed debris; (2) groundwater; and (3)
lagoons, ponds, drums and tanks. Using a current list of 34 tech-
nologies, the expert system recommends a range of response actions
for each contaminated area. Each response action typically includes
between three and 10 technologies.
  The cost system is independent of the expert system and can be
used after the remedial alternatives for the site have been selected,
using either the expert system or other means. The cost system
currently consists of 34 response  action technologies  including
containment, conventional treatment and alternative treatment. The
inputs, cost algorithms and outputs for each technology are con-
tained in separate cost modules which can  be used as  blocks to
build a complete response action  scenario. The user supplies a
limited number of scope-defining  parameters for the selected
response options (e.g., flow rates, treatment volumes, average tem-
perature during construction, etc.). The cost modules then develop
order-of-magnitude cost estimates for  the  options  using cost-
estimating algorithms.
  The CORA system is intended to be used during the RI stage
of the cleanup process.  Site data are limited at this point in the
process, and cleanup criteria may not be known.  The system is
designed to  provide preliminary cost estimates for a range of
response actions that are likely to be considered at  sites that have
not completed the RI/FS process at the time that the estimate is
required.
  The system is not intended to recommend one specific remedial
action. That decision is made by the states and the U.S. EPA after
analyzing additional technical, cost, policy and other site-specific
issues. The CORA system is intended to provide the user with a
set of reasonable response actions and cost estimates for each based
on the available site data. Consideration of a range of options is
appropriate, since the data and analyses required to recommend
a single option are not available. Thus, while the system is not
intended to replace the technology evaluation  process of an FS,
information provided by the model can be helpful in scoping sub-
sequent FS activities.

BACKGROUND
  During the early years of the Superfund program (1981-1983)
little historical data existed for developing cost estimates for the
wide variety of conditions found at Superfund sites. During this
period the program relied on "average" pricing factors to develop
376    SITE REMEDIATION

-------
budgets. The subjective nature of these pricing factors and the
absence of studies to confirm the factors were considered weak-
nesses in the program.
  In mid-1983, the U.S. EPA commissioned a study to attempt
to quantitatively define  pricing  factors for  remedial  actions.
Because  of scant historical  construction  cost information,  a
modeling approach for developing pricing factors was selected. In-
formation was obtained about site conditions at a small sample
of Superfund sites and a set of written decision rules was applied
to determine remedy selection.  The  sites in the study were
segregated into site types (e.g., landfills, drum sites) and the costs
of the remedies were estimated using a unit-pricing approach. The
resultant estimated costs were averaged and extrapolated to include
the 546 sites on the NPL. The estimates were then aggregated to
arrive at an average cost of construction for an NPL site.
  As the 1985 update of the FY 1986 budget approached, the U.S.
EPA sought to develop site-specific budgeting. The  U.S. EPA
attempted to disaggregate the 1985 results for use on a site-specific
basis. Efforts to refine these estimates pointed out the need for
a more accurate individual site-pricing technique. The CORA model
was designed to replace these earlier cost-estimating methods and
is now being used to estimate the  cost of outyear  Superfund
remedial actions on  a  site-specific basis. Cost estimates are
aggregated from the CORA model results and a variety of other
sources including RODs and FSs to form the outyear budget.
CORA SYSTEM COMPONENTS
Expert System
   The CORA model was built within the expert system shell. The
expert system might more accurately be described as a knowledge
base formed of decision rules for applying 34 representative con-
ventional and alternative technologies at  Superfund sites. The
decision rules reflect both engineering expertise and approaches
drawn from hazardous waste projects  and policy issues. Prefer-
ences for the selection of remedies with characteristics  established
in SARA are incorporated within the decision rules, as are emerging
policies and the U.S. EPA interpretations of the SARA language.
   The expert system analyzes a site by focusing on each contami-
nated area within the waste matrices. Separate decision rules were
developed for three waste matrices: (1) soils, sediment and mixed
 debris; (2) groundwater; and (3) lagoons, ponds, drums and tanks.
   The user responds to system-generated questions for each con-
 taminated area within a waste matrix being examined. The user
 can change his or her responses to questions posed by the system
 but cannot alter the decision rules. The system is intended to be
 run repeatedly to develop several sets of management options for
 each site. By examining alternative lines of reasoning leading from
 particularly difficult questions,  the user  generates a range of
 reasonable  response actions. The user can then assemble various
 response action  scenarios for costing  from the expert system's
 recommendations.  An example  of this process is  depicted  in
 Figure 1.
      Expert System Questions
          (Condensed)

   Are groundwater contamination levels
      above health—based limits

   What is the aquifer classification

   Is a POTW in  the vicinity

   Can permission be obtained to discharge
      to the POTW

   Is there a surface water in the vicinity

   Is metals removaf required

   Scenario  1  - To POTW
    Groundwoter extraction

    Metals precipitation

    Dlscorge to  POTW
       Initial
      Answers
        Yes

        Yes
         Yes

         Yes
Secondary
 Answers
   Yes

   Yes
Scenario 2 - To Surface Water
 Groundwater extraction

 Metals precipitation

 Ion exchange

 Discharge to surface water
                             Figure 1
                Example of Development of Scenarios
                                     Table 1 includes a list of the waste matrices and the current 34
                                   technologies that are candidates to be recommended by the expert
                                   system and costed using the cost modules. Each technology was
                                   selected based on its history of use and the ability to develop a
                                   cost estimate for it. Options that are either unproven in actual use
                                   or not well enough defined for cost estimates to be developed were
                                   not chosen. While this means that undemonstrated  technologies
                                   were not included in the model because of the uncertainty of cost
                                   estimates, the framework developed for current options facilitates
                                   the inclusion of additional  options.
                                                               Table 1
                                                 Cora Waste Matrices and Cost Modules
                                       WASTE MATRICES

                                        Sell), sediments,
                                         uncontolnid sludgs,
                                         ond mlKld d.brl,
                                        Logoons, ponds,

                                         tanks, ond

                                         drums
                                                                  TECHNOLOGY COST MODULES
Contolnment
Technologies
Cop (3 vortetl.i)
Surfoct) Contrail
Slurry Wall
Solldlllcollon
Removal
Technologies

Drum Romo*at
Soil Cicavatlan
Liquid ond Sludge
Removal
Croundwol.r extraction
Aettve V.nl Cat

CalLetton


Treatment
Technologies
Air Stripping
Vapor Pha.e Carton
Actlvatod Carbon
Motale Proclpltattan
ActNatod Sludgo
Son Vapor extraction

Son riuihlng
Homo Carton Unllo
Ollillo RCRA Troalm.nl
Offilla RCRA Incineration
Onille mclnoratlan

flare



Disposal
Technologies
Orftllo RCRA Landfill
Onitlo RCRA Landltll
Olftlle Solid Watt*
Landltll
Mieharge lป POTW
Dlicnarge lป Surlow
Wol.r

Miscellaneous
Technologies
Tronipoftellon
yunlclpal Water

Supply
Croundooter
Uonltertno
Aeceie Reilrlettoni
                                    The goal was to provide a system that can address the majority
                                    of sites; the system is not intended to address every site. "Out-
                                    liers" include sites with radioactive waste and mining sites. Situa-
                                    tions that  are  unique  and site-specific, however,  cannot  be
                                    anticipated and resolved within the scope of this system.

                                    Cost System
                                      The CORA cost system is used to develop order-of-magnitude
                                    cost estimates for sites after the response action scenarios are
                                    determined. The cost system currently comprises 34 cost modules
                                    and a system designed to organize the cost estimates by site, oper-
                                    able unit and scenario.
                                      Cost modules were developed following order-of-magnitude cost
                                    estimating techniques typically used to obtain FS cost estimates.
                                    Both capital and annual O&M costs are estimated. Items included
                                    in the cost module development are shown  in Table 2.

                                                               Table 2
                                               Items Considered in Cost Module Development

                                                      Site Remedial Action Cost Categories

                                                 Capital Costs                       O&M Costs
Technology
Conilrucllon
RA equipment
Tettlng/monHarlng
Dlipoiel
Troniporiallon
Cenlractor labor
•4ulpm.nl end
Temporary building.
P.rmon.nl bulldlngt
Temporary utlllllei
Permanonl ullllllii
Intereil and Inflation
during conilruellon
Site
Development
Ace. 11 control
Sll. preporallon
lor .qulpm.nl ond
building*
CLarlna and grubbing
Ulllltr eonn.ellont
Ulllltr r.locot1on*






Indirect
Casli
Slarlup cott*
Bid eonllng.net**
Scope conllngincle*
Blddtng and contract
admtnUtrallon
Con..r.,ctl.n monag.rn.nl
Chang, ardor nigollallont





Technology Indirect
0 4 U Costs
Supervlilon Taie*
Heallh and talety Ucini.i ond
yaterial* permit!
Utlllilii Etcro* lor
Tilling and rep1oe.rn.nl
monitoring Admlnl.trollon
and dlipoiel





                       The estimating techniques typically used are described in the fol-
                     lowing example.  An  engineer  with considerable experience  in
                     designing the technology defines the limits of the applicable range

                                                    SITE REMEDIATION    377

-------
for the design, develops the design scope variations across the range
for the design and obtains price quotes for major equipment (if
relevant). The cost estimator adds construction-related items and
translates the information into  a costing spreadsheet. The cost
estimator then performs a sensitivity analysis for each of the line
items to identify those that affect costs by more than 30% over
the range of the design. Cost algorithms are then developed for
each of the significant parameters. These significant cost-driving
parameters ultimately become the input parameters for the cost
module.
  The cost system is organized by site, operable unit, scenario and
technology. An example of how a site might be analyzed is shown
in Figure 2. The system  first asks the user to either specify a site
existing within the site cost data-base or designate a new site. The
user then designates operable units and scenarios to be considered.
Technologies for the scenarios are selected, and the cost modules
are run.  The user inputs the required costing parameters  and the
system calculates the capital and O&M cost estimates.

              SITE - SUPERFUND SITE  ANYWHERE. USA
    OPERABLE UNIT  1 - SOILS
    SCENARIO 1  - INCINERATION
     TECHNOLOGIES
           EXCAVATION
           TRANSPORTATION
           OFTSITC INCINERATION


    SCENARIO 1  ~ CONTAINMENT
      TECHNOLOGIES
           RCRA CAP
OPERABLE UNIT 2   GROUNDWATER

   SCENARIO I   SURFACE DISCHARGE

    TECHNOLOGIES

          GROUNDWATER EXTRACTION
          UETALS PRECIPITATION
          ION EXCHANGE

          DISCHARGE TO SURFACE WATER

   SCENARIO 2 - DISCHARGE TO POTW

    TECHNOLOGIES
          GROUNDWATER EXTRACTION

          METALS PRECIPITATION
          DISCHARGE TO POTW
                             Figure 2
                    Example of Site Organization
  Most cost modules provide the user with base-case default values
for some parameters.  By providing base-case values as defaults,
the user is given an opportunity to use and input known informa-
tion but is not prevented from developing a cost estimate if the
exact value is unknown. One example of this is the afterburner
temperature for an on-site rotary kiln incinerator. The required
fuel costs are  greatly influenced by the afterburner temperature;
however, not  many users know what will be required. In a case
such as this, the system provides a base-case value depending on
the contaminants of concern. If, at a later date, the user defines
this requirement, he or she can easily edit the input value and up-
date the cost  estimate.
  Outputs from the cost system are saved into a data-base for sub-
sequent analysis or updating and, in addition, are printed for a
hard-copy  record. Printed output is provided for individual tech-
nologies and for the entire site.  Examples of both types of output
are provided  in Figures 3 and 4.

EXPERIENCE
  In May  1987, cost estimates  were developed  using the CORA
model for 97 sites considered to be possible FY 1989 remedial action
candidates. This activity was accomplished with a team of the U.S.
EPA and CH2M HILL personnel.  Each team member worked one-
on-one with Regional Site Project Officers (RSPOs) to facilitate
the use of the CORA system.  The RSPOs contributed  varying
degrees of personal knowledge about the  site. This first-hand
knowledge was supplemented from other sources including the site
investigation  (SI),  the hazard ranking system  (MRS)  scoring
package, the workplan and/or the draft RI. Typically, two response
action scenarios were  estimated to bracket the cost range for the
site.
                                                                                   CORA CROUNOUATER COLLECTION COST MODULE
                                         SITE NAME:       SUPERPUND SITE, ANYWHERE USA
                                         OPERABLE UNIT;   CROUNOUATCII
                                         ESTIMATED START: LATE fy  19lซ
                                         SCENARIO:       DISCHARGE TO POTW
INPUTS
Parameter

Value
RESULTS
Component

Total
                                         No.  of collectlon/ซซt. uellt      (
                                         Averaqe veil depth lit|           10
                                         Lonaeat site dimension (ft)      100
                                         Averaa* tee>p {d*9r**B f)          SB
                                         Confidence level                M
                                         Protection level                C
                                CM ITU. COST
                                CONTINGENCIES
                                OIK COSTS
  12,0110
  10,000
     940
                                BYPRODUCTS FOR TRANSPORT/DISPOSAL:
                                                                             SOLIDS ICt)
                                                                             LIQUIDS ICALLOUSI
                                                          7
                                                      4.320
                                             Note for Croundvater Collection!

                                             The plplno. layout for groundwater collection  !• highly Kit*
                                             epeclflc.  Thlซ eodul* a*euMl the linear roouo,c of pip*
                                             equivalent to the one-halt the lonqcet lit* dlMMion.
                                             Coat* arc baaed on I* dl***t*r ductile iron pipe, burled vita
                                             4 ft of cover.

                                                                   Figure 3
                                                            Technology Cost Output
             CAPITAL COST OEVCLOPnort
SITE HAKE  SUPCKniHO SITE. ANYWHERE USA

INDIVIDUAL TECHNOLOGY COSTS
                                      BCC1OU  01
OPERABLE UNIT: SOILS

SCENARIO;  INCINERATION

EXCAVATION Or SOLIDS
OrFSITE RCRA INCINERATION
TRANSPORTATION TO OFfSITC RCRA INCINERATION

OPERABLE OMIT: GROUNOWATER

SCENARIO:  DISCHARGE TO POTW

NELL POINTS/OH COLLECTION
PRECIP/METALS TREATMENT STSTEM
DISCHARGE TO POTW
                                           SITE CLEARING COSTS
                                           ACCESS ROADS COSTS
                                                        SUBTOTAL

                                           GENERAL CONDITIONS
                                           COMMUNITY RELATIONS
                                           LOCAL BUILDING PERMITS
                                           START-UP COSTS
                                                          CONSTRUCTION SUBTOTAL
                                           BID CONTINGENCIES
                                           SCOPE CONTINGENCIES
                                                          CONSTRUCTION TOTAL
                                           PERMITTING AND LEGAL COSTS
                                           SERVICES DOB ING CONSTRUCTION
                                                          TOTAL CAPITAL COST
  210,000
I. BOO.000
  270.000
                                                                                                 42,000
                                                                                              1,000,000
                                                                                                 55.000

                                                                                              6,377.000
                                                       I, COO
                                                      SI,000
                                                                                                 55,600
                                                     110,000
                                                     130.000
                                                     120,000

                                                   7,011,600

                                                   1.400,000
                                                   1.100,000

                                                   5.512.600

                                                     470,000
                                                     950.000

                                                   10,932,(00
                                                                   Figure 4
                                                            Site Capital Cost Output

                                          Results from the CORA modeling effort were combined with
                                        cost information from other sources to develop the FY 1989 budget.
                                        A number of different analyses have been conducted on the data-
                                        base and the results of these analyses have helped the U.S. EPA
                                        shape thinking on the selection of remedy process under SARA.
                                        Figure 5 shows the capital distribution for modeled sites.
                                          Many other uses were discovered for the model during recent
                                        data collection efforts. CORA can be used to:

                                        • Screen possible remedies during early site studies
                                        • Provide a comprehensive analysis of the entire site
                                        • Evaluate potential data gaps in the remedial investigation
378    SITE REMEDIATION

-------
                              10 Sites
                               45.1%
53  Sites
  8.6%
      18 Sites
       16.3%
   NoU: Piretntog* of Tolol

       Capitol }* Shown
9 Sites
 19.0%
                           7 Sites
                            11.1%
                         Figure 5
             Capital Distribution for Modelled Sites

• Give order of magnitude estimates on the  potential  size of
  annual and multi-year State cost sharing requirements
• Provide an analytical means of estimating the average cost of
  NPL cleanup for all NPL sites

CORA MODEL VALIDATION
  Model validation is directed to the two components of the CORA
model: the expert system and the cost system.
  Validation of the expert system is being accomplished by peer
review. Independent experts have reviewed the decision rules and
found them  to be an acceptable representation of both SARA
requirements and common engineering practices in hazardous waste
cleanup activities.
  Validation of the cost system will be accomplished through the
"real world" validation procedure of comparing CORA model
estimates with ongoing or completed construction actions. The cost
modules will be calibrated using this technique.

CORA PHASE 2
  The Phase 1 CORA system was designed to be used by a team
of hazardous waste experts trained in the operation of the system,
working along with RSPOs. The experts provided both operational
help and  technical  interpretations of  the  system-generated
questions. The goal of Phase 2 is to enhance the CORA system
and the documentation such that the model can be used by RSPOs
without direct assistance.
  Development of the Phase 2 model is under way. The focuses
of Phase 2 work are to refine the cost modules and decision rules
and make the model more user friendly by including help screens.
A user's manual  and technical documentation also are being
developed. The system is being enhanced by addressing many of
the limitations of the Phase 1 system. Enhancements include rules-
of-thumb for estimating scope parameters, the addition of 10 cost
modules and the expansion of design ranges for several of the tech-
nologies. The Phase 2 system will also include an easy method to
update cost information;  a cost indexing system  is being incor-
porated into each cost module to facilitate periodic updating.
  The U.S. EPA plans to make the CORA model available to each
U.S. EPA regional office early in calendar year 1988. The approach
is for a member of the regional remedial staff to serve as the
primary operator of the model for the region. Creating a formal
locus of responsibility for operating the model ensures that one
person will become familiar with the model systems and operating
environment and can act as a general resource.
  While the primary operator might assist each RSPO during the
initial trial, the intent is to provide a system that is both useful
and straightforward to  use.
  Current intentions  are  to use the CORA model to assist in
developing outyear budget estimates annually.
                                                                                               SITE REMEDIATION    379

-------
                     Heavy  Metals-Contaminated  Soils  Treatment

                                              Peter S.  Puglionesi,  P.E.
                                                   Jaisimha Kesari
                                              Michael H. Corbin, P.E.,
                                                 Roy  F. Weston,  Inc.
                                             West Chester, Pennsylvania
                                                  Erik B.  Hangeland
                              U.S.  Army  Toxic and Hazardous Materials Agency
                                      Aberdeen Proving Grounds,  Maryland
 ABSTRACT
   Metals contaminated soils are a problem at many remedial action
 sites. This study identified existing, developmental and conceptual
 options for treatment of heavy metals contaminated soils and
 residues in order to evaluate treatment technologies which may be
 candidates for further research and development. WESTON per-
 formed this study under contract to the U.S. Army Toxic and
 Hazardous Materials Agency (USATHAMA).
  The first part of this study involved the characterization of soil
metals concentrations at various U.S. Army sites and a review of
regulations/guidelines/criteria for residual soil metals concentra-
tions applicable to contaminated soil cleanup. These characteri-
zations indicated that several Army installations had elevated total
metals concentrations in their soils and that chromium, cadmium
and lead were the most prevalent metals.

  A detailed literature search identified 21 potentially viable treat-
ment technologies.  Process descriptions were prepared which
included flow diagrams, use backgrounds, and expected levels of
performance. These technologies were subjected to a preliminary
feasibility screening. The following technologies were selected for
more detailed evaluation:

• Microencapsulation
• Roasting
• Extraction (on-site)

   Preliminary  concept  designs including process  flow sheets,
 material balances and equipment sizings were developed for these
 technologies. In addition, total project  costs were estimated for
 each technology. The three technologies were then subjected to a
 detailed technical evaluation.  Roasting was identified as the best
 technology with respect to expected high performance and relia-
 bility, competitive cost and probability of successful development.
 Other technologies which were highly rated were not retained for
 consideration because they were either commercialized or the
 subject of other USATHAMA studies.

 INTRODUCTION
   Treatment of heavy metals contaminated residues and soils is
 one of the most difficult challenges in the environmental field
 because metals cannot  be biodegraded or destroyed. Soils and
 sludges containing heavy metals are an increasing concern across
 the country.
   The primary method historically used for management of heavy
 metals contaminated soils and  residues has been land disposal. The
 high costs and risks associated with land disposal and future RCRA
land Till prohibitions necessitate the exploration of environmentally
sound alternatives to land disposal.
  The  U.S.  Army s industrial operations  such  as equipment
rebuilding and repair, munitions manufacturing and  disposal
involve heavy metals contaminated residues. In an effort to develop
more environmentally sound options for the treatment of these
heavy metals-contaminated soils and residues, the U.S. Army Toxic
and Hazardous Materials Agency (USATHAMA) initiated  an
evaluation of treatment technologies which may be candidates for
further development. WESTON, under contract to USATHAMA,
performed an in-depth engineering study and analysis to identify
and evaluate the most promising technologies for treatment of
heavy metals contaminated soils.

DISCUSSION
  The study commenced in early 1986 with  a  review of existing
data on the metals concentrations found in soils at various U.S.
Army installations.  Based  on  total  metals  concentrations, the
existing data indicated that chromium (Cr), cadmium (Cd) and lead
(Pb) are the metal species most often present at elevated  levels.
Therefore, the evaluation of treatment technologies was based on
soil containing these metals.
  State and Federal regulatory requirements for contaminated soil
remediation were reviewed to identify and  target soil metals con-
centrations which should be achieved by a remedial treatment
process. This  review indicated that specific treatment objectives
for metals contaminated  soils currently do not exist. Treatment
objectives typically are based on site-specific evaluations of migra-
tion potential and assessments of potential impact. Based on current
hazardous waste disposal regulations,  the  achievement of EP
Toxicity limits is considered a minimum treatment target, and the
achievement of background levels is considered a maximum  target.
Based on this regulatory  framework, it was determined that any
technology selected  for possible  development  should,  at  a
minimum, be capable of treating the metals contaminated soils to
reduce their teachability  below EP Toxicity limits.
  An extensive literature search was conducted and field researchers
were contacted in order to identify potentially viable metals con-
taminated soil treatment technologies. Many  technologies were
investigated which have not typically been applied to soil treat-
ment or site remediation  but which potentially could be adapted
for this type of application. Several  computerized data bases
including  NTIS, Pollution Abstracts and papers presented at
engineering meetings were accessed to retrieve pertinent literature.
  Personal contacts included those with the following persons/
organizations:
 380    CONTAMINATED SOIL TREATMENT

-------
• U.S. EPA Research Project Officers
• DOD Research Project Officers
• University researchers
• Commercial enterprises
   This search identified 21 treatment technologies. Most of these
technologies are new, innovative methods for treatment of metals-
contaminated soils.


  Since  metals  cannot  be readily destroyed  as  can  organic
pollutants, the approaches to metals "treatment" primarily include:


•  Immobilization which integrates metals into  a stable matrix
•  Removal of metals from the soil and concentration to form a
   smaller volume of waste residue
•  Isolation of the soil to prevent its contact with the environment
   The technologies identified are listed in Table 1 along with the
 approach to metals treatment emploYed (e.g., immobilization,
 removal, etc.) and the type of treatment employed (e.g., chemical,
 biological, etc.) Certain technologies are classified as "on-site" and
 "in site". An  on-site technology is one where the contaminated
 soil is excavated and treated, while an in-site technology is one
 where the contaminated soil is treated in place (i.e., no excavation
 is required).
   A technology profile and concept engineering process sketch were
 prepared,  and the technologies were evaluated and compared using
 criteria pertaining to USATHAMA development/implementation
 requirements. The primary objective of this evaluation was to
 identify those  technologies  which had a  high probability of
 achieving the treatment goals and of being successfully developed,
 scaled up and  applied  in  USATHAMA  remediation  effort
 programs. The following  development/implementation  criteria
 were used in this evaluation:
 • Treatment effectiveness—Ability to remove metals or render the
   substrate nonhazardous (Per EP Toxicity criteria) to meet regu-
   latory or  cleanup objectives—Rating Factor  =  4
 • Long-term stability/performance—Assess the permanence of the
   treatment performance in the long-term. The effectiveness of
   the treatment process in rendering the soil nonhazardous may
   be  impacted  by  long-term environmental conditions  (e.g.,
   weathering, infiltration, PH, etc.)—Rating Factor =  3
 • Residual  treatment/disposal  requirements—  If  potentially
   hazardous residuals  may  be   formed,  additional treatment
   processes  or secure disposal will be required. If residuals treat-
   ment is not addressed, it could require further development. If
   residuals disposal is necessary, this negatively impacts future risk
   and liability—Rating Factor = 3
 • Flexibility—Ability to treat various soil/site types, to treat other
   waste streams, to treat for organics and metals, to be linked to
   other organic treatment processes or to handle materials with
   variable physical consistency—Rating Factor  = 2
 • Material throughput rate—Ability to process large quantities of
   soil or anticipated abilitY to scale up to meet this objective-
   Rating Factor =  2
 • Potential disqualifiers—Evaluate potential fatal  flaws to deter-
   mine if any would prevent technology development or imple-
   mentation (a  non-numerical consideration)

   These criteria were selected as the most important criteria which
 address technical feasibility (i.e., applicability and achievement of
 treatment and project goals).
   The above criteria were given different weighting based on their
 relative importance with respect to USATHAMA remedial objec-
                              Table 1
               Technologies Identified for Treatment of
                     Metals Contaminated Soils
                                                                                              Treatment-Approach
                                                                                                                Development
   On-iieซ plaau ซrc      TherMl-1
   (w/aetal recovery)      Removal
                    ThetMl-lBMobiliiatlon
                    ftieraul-lBKJbl Illation
                                                                            High gradient magnetic   Physical-Removal
                                                                            •eparation
   In situ preclpil
   by vapor phase
   application
                                                                            Vegetative uptake      Biological
                                      Pllot-Bcale
   Geologic Isolatic

   Secure landfill
                    Phyaical-Iaolatioi
   On-tIt* ion exchange
                                                  into tht ground to heat
                                                  th* ioil which forma a
                                                  factoring ovtn li u*e(
                                                  to heat and glaailfy
High enargy electric ai
destroys organic* and
glaaaKles or ซj*•!(!ซ•
netala foe subsequent
                                                  Radii
01 gaslfUa twtala (or
subsequent condensation.

Additives and h**t used
to convert aoll to a
atablt crystalline ot
glaaa matrix.

High magnetic gradient
uaed to deflect magnalic
or paranagnetlc par-
tlclea containing mttala
from thoae without
metals.
the chlorld* c.sulta In

sequent condensation.

PaaBlbl* adaptation of
waatewatec precipita-
tion techniques to Iranto-
blllia metala In exca-
vated aoll In continu-
ous flow process equip-
ment.

Possible adaptation of
vastewater precipitation
technique* applied

In place.

Sana as above except
UBป gsacoua preclpi-
tants.

Chelatlng agents or
                                                  mobillie Mtala. An
                                                  aasoclated metala recov-

                                                  nology Is requited,

                                                  Sane •* above except
                                                  applied In situ with
                                                  solution recovered via
                                                  gtoundwatซr.

                                                  Metals accumulate In

                                                  not directly addressed.
                                                  Chemical fixation In a

                                                  mixture coating wtth a
                                                  low permeability
                                                  mixture.
                                                  Coating with a low
                                                  permeability mixture.
                                                  soil with a low perr
                                                  ability material.

                                                  Ion exchange materials
                                                  incorporated Into aoil.
                                                  Transfer of metala frซ
                                                  aoil to Ion exchange
                                                  covery/concentration
                                                  technology 1> required.
lives. This is noted above as the "Rating Factor". In the technical
evaluation, numerical scores ranging from 0 to 5 (with 0 being least
desirable and 5 being most desirable) were  assigned for  each
criterion for each technology. These scores were multiplied by the
Rating Factors and the products summed to yield a total score for
each technology. These evaluation were reviewed by a peer group
consisting of engineers and scientists to finalize the  scores for each
technology.
  The following is a tabulation of the top 10 technologies based
on their consensus scores:
                                                                                        CONTAMINATED SOIL TREATMENT     381

-------
          Rank
Technology
Score
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
On-site vitrification
Microencapsulation
Roasting
Stabilization (admixing)
In situ vitrification
Geologic isolation
Secure landfill
Chloride volatilization
Microencapsu latio n
Extraction (on-site)
57.5
55
54.5
54
51
49
48
44.5
41
40
   Preliminary consensus rankings were discussed at the meeting.
 During the course of this discussion,  it was recommended that geo-
 logic isolation and secure landfill should not be considered further
 because they were not treatment "technologies" per se. Macro-
 encapsulation was ruled out because of fatal flaws which included
 high costs and institutional concerns  (e.g., soil is not rendered non-
 hazardous by macroencapsulation alone, and long-term contain-
 ment and monitoring must be provided).
   It was suggested that although roasting was far more attractive
 than chloride volatilization, they might be combined and studied
 in a single technology development program. However,  further
 technology evaluation and cost estimation should focus on roasting,
 the most promising technology.
   It was suggested that microencapsulation could be combined with
 stabilization (admixing) and studied as one technology. Stabiliza-
 tion (admixing) has become extensively commercialized in recent
 Years, however, and its application to a remedial action site can
 be evaluated on the basis of bench-testing and historical informa-
 tion for each application. Because it is considered a proven tech-
 nology with a substantial performance track record,  additional
 R&D for the stabilization technology was not considered to be justi-
 fied. Therefore, further evaluation and cost estimation should focus
 on microencapsulation, the technology which is more likely to
 require further development  effort  prior to implementation.
   USATHAMA indicated that both vitrification technologies (in
 situ and on-site) were already being  studied as part of the installa-
 tion cleanup program. Although these were highly rated, the need
 for a  separate R<6D program for these  technologies was not
 considered justified at this time considering the ongoing work. Con-
 sequently, extraction (on-site) was  recommended as a  candidate
 for further study despite its having a lower rating than the vitrifi-
 cation technologies. Extraction (on-site) was rated lower primarily
 on the basis of the uncertainty of achieving the required treatment
 performance and the need for subsequent processing steps for the
 extract solution. However, the potential for efficient, cost-effective
 processing of metals-contaminated soils  made  this technology
 worthy of further evaluation.
   Thus, the outcome of this preliminary technical evaluation was
 the selection of the  following three technologies for further in-
 vestigation:

• Microencapsulation
  This technology involves the  mixing and heating of metals-
contaminated soils with a polymer  or other thermoplastic (such
as asphalt) in an extruder at moderately high temperature.  Metal
contaminants in the soil are immobilized in the thermoplastic matrix
to prevent their release to the environment. Based on its low leach-
ability, the treatment product could be backfilled or be disposed
as non-hazardous waste.

• Roasting
  Contaminated soil and additives,  such as Kaolin, are heated in
a rotary kiln or a multiple-hearth furnace. The anticipated operating
temperature is approximately two-thirds of the metal  s melting
temperature. Metals in the product are immobilized in a ceramic-

382     CONTAMINATED SOIL TREATMENT
 like matrix. The product is expected to exhibit extremely low leach-
 ability and to remain stable in the environment. The treatment
 product could be backfilled or be disposed as non-hazardous waste.
 • Extraction (on-site)
 Metals are removed from contaminated soil by using appropriate
 extracting solutions which chelate and/or dissolve the metals. The
 extractant is separated from the soil and further treated to either
 recover or concentrate the metals. The concentrated metals stream
 (typically chemical  precipitation  sludge)  can  be disposed as
 hazardous waste, and the treated soil can be backfilled on-site.
  Preliminary concept  designs  including process  flow  sheets,
material  balances, major equipment sizings and estimated costs
were developed for each of these three selected technologies based
on soil treatment for a "generic" installation with Cd,  Cr and Pb
contaminated soil. Preliminary process block flow diagrams are
presented in Figures 1,2 and 3, and the corresponding process flow
sheets are shown  in  Figures 4,  5 and 6.
                                                                 Figure 1
                                              Block Process Flow Diagram for Microencapsulation
                                                                 Figure 2
                                                  Block Process Flow Diagram for Roasting
                                                                 Figure 3
                                                 Block Process Flow Diagram for Extraction

-------
                    Conwyw   ScrMHK)
                                             Conuirwn for   Truck
                             Figure 4
              Process Flow Sheet for Microencapsulation
    Convoyor /

      ScrHned
                            Figure 5
                 Process Flow Sheet for Roasting
      Jsa-i
ElCMM ^
                             Figure 6
                  Process Flow Sheet for Extraction
   The following assumptions/conditions were used in developing
 and comparing the concept designs for the three technologies:

 •  Two separate quantities of soil to be processed, representing
   a range of sites, would be investigated as follows:
   — Site  1 - 10,000 tons
   — Site 2 - 100,000 tons
• Soil type: Silty clay with sand

• Soil density prior to excavation:
  Average 100 lb/ft3
  Range 90 to 110 lb/ft3

• Soil bulk density after excavation:
  Average 75 lb/ft3
  Range 65 to 85 lb/ft3

• Soil moisture content:
  Average—20%
  Range—15 to 30%

• Soil is assumed to be contaminated only with the following
  metals at the indicated concentrations:
                                                                                           Concentration (as total metal, mg/kg)
Range
Average
Range
Average
Range
Average
1-3,000
1,500
1-5,000
2,500
0-500
250
  Metal

Chromium


Lead


Cadmium
• Time required for cleanup to be completed was assumed to be
  1 year. This was applicable to both quantities of soil (i.e., 10,000
  and  100,000 tons)
• While a mobile unit may be most desirable, it would be difficult
  to cost all three technologies on this basis given the varying levels
  of development. For comparison, costs were developed based
  on a permanently installed on-site process unit. However, the
  potential for design of a mobile or reuseable/relocatable process
  unit  also was assessed.
• Potential applications of soil treatment  for metals may also
  require treatment for volatile organics or explosives at the same
  time. Likewise, these technologies potentially could be applied
  to metals bearing sludges. Since the principal purpose of the
  evaluation is to compare technologies prior to development,
  evaluations of subalternatives and options were limited. There-
  fore, the primary basis for concept design and technology com-
  parison was an application with only metal contaminants. Some
  consideration was given to alternative process configurations and
  treatment costs on a less rigorous basis.
• Assumed labor  and energy  rates were used  to  calculate the
  operating costs for the three technologies. These rates represent
  national averages applicable to the hypothetical site considered
  in the designs. Since the total project time  was  1 year, the
  capital  and  operating costs were added  to obtain  the  total
  annual project cost.

  The preliminary concept designs were subjected to a second-level
detailed evaluation using numerous technical feasibility, develop-
ment and implementation criteria  and the estimated costs (in
relation to those for the conventional technique of excavation and
off-site disposal). The objective of this  evaluation was to deter-
mine which of these three technologies should be candidates for
further research and development. This second evaluation was done
on a relative basis with scores from 0 to 3 (0 being least desirable
and 3 being most desirable) assigned to each criterion for each tech-
nology. Again, these scores were multiplied bY the rating factors
and summed to yield the total numerical score for each technology.
Table  2 summarizes  the comparative numerical rankings for the
three technologies. Projected costs for applying each technology
at sites containing 10,000 and 100,000 tons of metals-contaminated
soil are presented in Table 2. Table 3 summarizes the advantages
and disadvantages for each technology based  on  the detailed
evaluation.
                                                                                 CONTAMINATED SOIL TREATMENT    383

-------
                                Table 2
            Comparative Rankings and Costs of Technologies
                                                                                                      Table 3
                                                                                   Summary of Individual Technology Evaluations
      Critorla/toehBologr
      J.  Loajf-tom
ซ.  Throughput
5.  Matorlala haodliof
t.  Air coatrolc

I.  Ka>o of oparatiooa

ซ.  Traupor (ability
10  ••liability
         Projoct coit
      TOTAL HUMCKICAL SCOM
      KซtiMato4 Total rrojoct
      Co.t  (1).
      10.0OO TOM or Soil

      100.000 TOM of Soil
                         S t.MO.OOO   S  >. TOO,000   f VOSO.OOO

                         S2V.C1S.OOO   I10.9SO.OOO   110,270.000
         1.  a*a*4 oa oa* r*ar o*ratloa of oporatioaa.
                                                                                                                  l*lซ Mn4II*t r*-   •*oeปiป aai aM MM
                                                                                                                  ซ*t> •(• MM*     applta* to tปป *(Mt-
                                                                                                                  i UkM UM *UMr   MM of Mt*lป enwal-
                                                                                                                  TTfinrliti lit ••     ซ*t*^ •ell*. U*orซt*rr
                                                                                                                       wit OMU
                                                                                                               ซMU Ml l
                                                                                                               klซR, tta (
                                                                                                    ipi (**4ป      (tewtt ii iii
                                                                                                    WMU ซf      Mปtl •*•
                                                                                                    t Vlik llltU   *lt* felflMf
                                                                                          •1 Mrf/ot Cปll-ซuli
                                                                                          C.pll.l 00. t. *l. klfk Wt     ttlfk.              (My M
                                                                                          •OMMlll Of MปU CM M                      I Ml*.
                                                                                          •Cftl***4l Mil CMtl Mซ      WtOt t|lM M* •••*•ซ
                                                                                          • lซMiriซซlปlv '•**** •*•-     ซMซ tftll IซซNCM IW
                                                                                          Mil r*MซMlซซ r*uซ Mซ     Mklllty •! t
                                                                                           • •!*• ••flซซ7 •( Ml I
CONCLUSIONS
   Roasting was found to offer the potential for high performance,
effective treatment through long-term immobilization of the metals
in a ceramic-like matrix and cost-effectiveness at high throughput
rates. Roasting is also a particularly attractive alternative where
incineration is being considered for soil which may contain organic
contaminants as well as metals. The performance of extraction is
less certain, as removal of metals in certain forms and certain types
of  soil may  prove to be difficult.  However,  if development  is
successful, it also offers the potential for cost-effective decontami-
nation of metals-contaminated soil. While microencapsulation is
expected to be effective for immobilization of metals, very high
operating costs are anticipated  due to high raw material costs.
Roasting is the technology recommended for further research and
development efforts.
                                                                                    •ป*•! *( Mt>l* IfOB Mill.    M lซปปtM I

                                                                                    f**tซซ(tMt U4 Mill ซM-      *ซCk *ltซ-ซ lym* Of

                                                                                    MllfMM.               Mil* Mซ MUl CM-
                                                                                                                          tt*Ctlซl ซwซt k*



                                                                                                         CM! I* MซM4 !• lrซ*t  kMfMI MM* ป** ty)
                                                                                                         Mill .IU W.MICI.    M*t M MUปIMซ.
                                                                                                         rtป}*ci eovti •(• •*ป•!-
                                                                                                         tl*ซ to f*Mซปl ปซ<
                                                                                                         • lซ*fซ ซIWซ*ซ1 CWll
                                                                                                         ••.let. IM tซซ* •!• ซ••*•-
                                                                                                         ซMt a> Ulitil vail Mill
384     CONTAMINATED SOIL TREATMENT

-------
               A  New Method  To  Characterize  Contaminated  Soils

                                              Namunu J. Meegoda, Ph.D.
                                                  Prasanna Ratnaweera
                               Department of Civil and Environmental Engineering
                                         New Jersey Institute of Technology
ABSTRACT
  There are a large  number of hazardous waste  sites located
throughout the country, of which some are still not identified by
the authorities. In many existing landfills, failure of clay liners have
been reported, giving rise to further contamination of neighboring
areas as well as the groundwater. Therefore, it is very important
to characterize the contaminated soils to assess the danger caused
by such spills and to mitigate the damage.
  Gradation, Atterberg limit tests X-ray analysis and Differential
Thermal analysis are currently adopted to characterize the con-
taminated soils. These methods may not  be appropriate as they
do not show the physico-chemical interactions that occur in chem-
ically contaminated soils.
  The electrical properties such as conductivity and dielectric cons-
tant  of the soil and  the pore fluid show the physico-chemical
interactions exist in a chemically-contaminated soil. Therefore, the
characterization of chemically-contaminated soils using the elec-
trical method is presented in this paper. The electrical method
eliminates evaporation of the volatile pore fluid and segregation
of immiscible fluids. The proposed method does not require human
handling of toxic chemicals and hence is much superior to the cur-
rent  methods.  Since the  electrical  measurements  are  non-
destructive, field conditions may be maintained on structure sen-
sitive soils.

INTRODUCTION
  A large number of hazardous waste dump sites some identified
and some yet to be discovered, are located throughout the United
States Soils in and around hazardous waste dump sites may be con-
taminated due to seepage of chemicals. Contaminated soils are fre-
quently encountered  in  construction,  especially in heavily
industrialized areas. Therefore, identification and characterization
of contaminated soils, for engineering purposes, has created a new
dimension in geotechnical engineering.
  The engineering properties of a soil depend on the composite
effects of the compositional and environmental factors. The com-
positional factors such as mineralogy, pore fluid composition and
concentration determine the potential behavior while the environ-
mental factors such as orientation of the particles, stress state and
cementation determine its in-situ behavior. Therefore undisturbed
samples or in-situ measurements are required for proper charac-
terization of engineering properties of a  soil. Furthermore, the
analysis and interpretation of laboratory tests create additional
problems as the geotechnical engineers are presently knowledge-
able in  geotechnical properties of soils with water as pore fluid.
When the pore fluid consists of chemicals, the physical properties
such as vicosity, density and surface tension differ significantly
and such variation should be taken into consideration when inter-
preting results of geotechnical tests performed on soils contami-
nated with chemicals.
  Currently, the characterization of an uncontaminated soil con-
sists of obtaining representative field  samples and performing
simple index tests to identify and classify the soil in order to obtain
the soil's engineering properties. Presently, the Atterberg Limit tests
and gradation tests are used extensively for  such classifications.
The X-Ray diffraction and Differential Thermal Analysis (DTA)
tests are occasionally used as such indices to  identify and charac-
terize soil.
  Even though, X-Ray diffraction, DTA, and electron microscope
studies yield the mineralogical data, it is not  possible to quantify
the interaction of two or more minerals with pore fluids to predict
the clay structure.9 Gradation results may not be very useful for
fine-grained soils. The Atterberg limit test, which is extensively used
as an index for soil classification, is now being used as an index
to classify chemically contaminated soils. The results of intensive
research by Lambe  and Martin7"" for the purpose of utilizing
Atterberg Limits for compositional analysis clearly  shows the
inability of Atterberg Limits to account for the cementation and
interstratification of natural soil deposits. The inherent difficul-
ties such as repeatability, segregation of components, evaporation
losses and errors due to human handling make Atterberg  Limits
a questionable index for compacted soils in contaminated en-
vironments.
  For the purpose of characterization and identification of con-
taminated soils, it is necessary to understand the clay-pore fluid-
electrolyte interaction. The physico-chemical interaction in soils
may lead to a functional index that may identify and characterize
the contaminated soils.

PHYSICO-CHEMICAL INTERACTION IN SOILS
  The clay-pore fluid-electrolyte system produces a given clay struc-
ture. The clay structure is the result of net inter particle repulsion
due to the formation of the double layer  and the inter particle
attraction due to Van der Waal forces. The  amount of physico-
chemical interaction in  a  soil is expressed as the  Double Layer
Thickness6'5. The thickness of the double layer which is the dis-
tance between two clay particles depends on the dielectric cons-
tant of the medium, absolute temperature, electrolyte concentration
and the valence of the pore fluid electrolyte. The change in chemi-
cal composition and  concentration of the pore fluid  results  in
varying double layer thicknesses to produce  different clay struc-
tures. Different  clay  structures may show completely different
                                                                              CONTAMINATED SOIL TREATMENT    385

-------
mechanical properties for the same clay.
  The reduction in double layer thickness gives rise to a floccu-
lated structure while an increase produces a dispersed structure.
These  two clay structures have  been identified with  distinctly
different engineering properties such as swell and shrinkage, com-
pressibility, permeability and shear strength. It is interesting to note
that most of the natural and compacted soils which are contami-
nated or uncontaminated may not have ideally flocculated or dis-
persed structures,  but structures in-between those depending on
the amount of physico-chemical interaction present in the clay-pore
fluid-electrolyte system.
  The electrical properties of soils, such as dielectric constant and
conductivity, are found to vary with different soil structures for
the same soil. Therefore, the electrical method is used to identify
and characterize the contaminated soils.

THE ELECTRICAL METHOD
  The proposed electrical method measures the dielectric constant
and conductivity of soil in the  radio frequency range (10*-10* Hz).
These  measurements are made with the aid of capacitance  and
resistance of a soil sample. The apparent dielectric constant  and
conductivity of soils were found to vary  with frequency (Fig. 1).
With the increase  in input frequency from 1 MHz to  100 MHz,
the dielectric constant may decrease and the conductivity may rise.
This variation is termed the dielectric dispersion Ato, and is due
to composition and  heterogeneity of the soil-water-electrolyte
system2-'5. If there is no soil-water interaction as in sands the die-
lectric dispersion in radio frequency range will be zero.
                             Table 1
            Comparison of Current and Electrical Methods
                Of Characterizing Contaminated Soils
              2     34    t  I  10    16 20  ?6 JO 5540 40 6070(090
                        FREQUENCY In MHZ


                            Figure 1
              Variation of Dielectridc Constant, < and
                Conductivity, 6 with Input Frequency
                   (after Arulanandan et al, 1983)
   The dielectric dispersion varied with the type and amount of
 chemicals  in the pore  fluid15. The dielectric dispersion is very
 sensitive to the dielectric constant of the pore fluid, electrolyte con-
 centration, valence of the pore fluid electrolyte. Therefore, die-
 lectric dispersion was used  in  this  research  to  identify  and
 characterize  the contaminated  soils in  order to predict the
 engineering properties of contaminated soils. The distinct advan-
 tages of using the electrical method over the current methods is
 given in Table 1.
   At present, the electrical method is used to predict the in situ
 properties of uncontaminated soils such as permeability, compres-
 sibility and strength1'3'12.  In this research work, we attempted to
 predict the in situ engineering properties of contaminated soils.
   Another objective of this research was to quantify the soil struc-
 ture produced by a given chemical in order to study the feasibility
 of replacing the testing of soils contaminated with toxic chemicals
 by the testing of soils mixed with equivalent non-toxic chemicals.

BASIS
UT/MflXZS
DISACVMttXZS
CUKRBrr MEDBC6
Current Bethod* are baaed on
elaple tecta adopted during the
Infancy of the •clanoe. Tnow
Include relative demity,
etandard penetration teats,
Atterbero; lielta and
coBpaalticnal amlyei*. (X-rey,
Differential Thermal minim,
etc.)
Relatively BconceUcal.
1-Requlra hvutan handling (todo
•aterlal)
2.The eoil etructure U i1eelliijป1,
3.Evaporซtlonof Pore fluid.
4 frtijiegiitlai of li-mif-IM* MuHi
OGDUT.
jxBCBuai- MBDOO
Electrical properties of
(oil* reflect the phyaioo
chemical interaction that
i* present in the clay-
Pore fluid electrolyte
yatซป in aoila.
IJbpld.
2.RepeetAble.
l.Hcn-dectructlve, hence
lit-cltu Beaaumenta
are poeaible.
4. Principally furriaeBfltal.
1 nerjilre technical akllli
EXPERIMENTAL PROCEDURE
  A detailed experimental study is under way to investigate the
possibility of identifying and characterizing chemically contami-
nated soils by the electrical method and to interpret the geotechni-
cal test results. Three artificial soils namely Kaoline (Soil #1),  \5Vป
Montmorillonite and 85ro Kaoline (Soil  #2), Illite (Soil 1/3) and
two natural soils namely New Brunswick clay (Soil 14) and Yolo
Loam (Soil W) were selected in this investigation. Five chemicals,
glycerol, 1 -butanol, 1 -propanol, acetic acid and O.SN sodium chlo-
ride solution,  were used  in this  investigation. The geotechnical
properties of soils selected are  listed in Table 2; the physical and
chemical properties of the fluids are listed in Table  3.

                            Table 2
                 Geotechnical Properties of the Soils
TV?**
of
Soli
Soil 11
SOU 12
Soil 13
Soil 14
Soil 15
Liquid
Lifldt
%
54
85
64
36
35
Specific
Surface
Area (J/q)
80
150
105
115
65
                             Table 3
     Physical and Chemical Properties of the Fluids Used at 25 ฐC
Type of
tare Fluid
Nad -Water
Solution
Glycarol
1-Propanol
1-Butanol
Acetic Acid
Hoi.
Haloht

IB
92.11
•0.11
74.12
60.05
b.p.
ฐC

100
182
97.4
117.2!
117.9
Denalty
(9/ca3)

1
1.2613
0.8035
0.8098
1.0492
Dielectric
Ocnetant

80
42.5
20.1
17.1
6.24
Viaooalty
(cp)

o.a
1490
2.254
2.948
1.155
Surface
TMlon
(dyn/ca)

73

23 78

27
 386    CONTAMINATED SOIL TREATMENT

-------
  The electrical measurements, consolidation tests, compaction
tests and strength measurements were made on the above soils with
different chemical combinations of the pore fluids.
  The electrical measurements in horizontal and vertical directions
were  made using  two cells2'12. A Hewlett-Packard impedance
analyzer was  used  to measure  the  electrical resistance  and
capacitance of the soils in horizontal and vertical directions using
the above cells. The electrical measurements were made for all the
soils with distilled water as pore fluid. A few tests were made with
pore fluids other than water.
  The sample preparation procedure described in Meegoda12, was
used in  this research. Two samples of soil in two cells were con-
solidated to different pressures and the electrical measurements
were made. The electrical measurements were reduced to eliminate
the line and cell impedance  in order to obtain the dielectric dis-
persion data for each stage of loading in horizontal and vertical
directions. The magnitude of dielectric dispersion values for each
test was computed using the procedure outlined in Meegoda12.
The Compression Index for  the above tests were computed from
the load settlement data.
  Compaction tests were performed to obtain the variation of dry
density with fluid content (ratio of the weight of fluid to weight
of dry soil) for Soil #2 with different ratios of water to chemicals.
The strength tests for Soil #2 with distilled water showed that soil
compacted to the same dry density on wet and dry side had similar
residual shear strength. Therefore, subsequent shear strength  tests
were  performed only on wet side of compaction for Soil #2  with
different pore fluids.

EXPERIMENTAL RESULTS
  Three physical properties of contaminated soils, namely the com-
paction, strength and compression, were measured. The results for
each  geotechnical property is discussed under a separate heading
and is given below.

Compaction tests
  Figure 2 shows the compaction test results for Soil #2 with vari-
ous ratios of 1-propanol and water as pore fluids. The conven-
tional method of reporting  compaction test results may not be
suitable for contaminated soil because the physical properties of
pore  fluids are different from water.
                                                PROPANOL *  HATER I
                                                         00.0
                                                         75.0
                                                         so.o
                                                         00.0
       26.0    28.0     30.0    32.0     34.0

                Fluid Concent, wX
                            Figure 2
   Variation of Dry Density 7d with Fluid Content, w for Soil #2

   A new procedure was adopted which consider the density, (G*)
 and the volume fraction, V (the ratio of fluid volume to water
 volume) of the contaminating fluid. Both the density and volume
 fraction of contaminated fluids play a major role in compaction
characteristics of the soil. By modifying the fluid content w by
w*(V + 1)/(V + G*), a meaningful plot was obtained (Figure 3). One
may appreciate the compaction curves which were scattered  in
Figure 2 plots properly in Figure 3 with wet sides of all curves
plotting parallel to zero air void ratio curve. Furthermore, in Figure
3, the modified optimum fluid content moves to the right giving
a lower optimum dry density. This shift is attributed to the increase
in vicosity of the pore fluid. With the increase in pore fluid vis-
cosity, most of the energy input in the form of compactive effort
is lost  against friction producing a lower curve. We suggest that
if compaction test results are to be plotted for contaminated soils
it should be plotted as suggested above.
                  30.0     32.0

                Modified Fluid Conci
 34.0    36.0     38.0     40.0
/-uMV+1 )/(ปซ;*>, V-V / V  C*- O.B
                           Figure 3
 Variation of Dry Density, Xd with Modified Fluid Content for Soil #2

Strength tests
  Table 4 shows the unconfined compression test results for Soil
#2 with various pore fluids. The unconfined compression tests were
performed on samples compacted on wet side to a dry density of
1.30 ฑ 0.5 g/cm3. The results show nearly a two orders of mag-
nitude change in shear strength. The pore fluid viscosity is found
to be of major influence on the shear strength. Figure 4 shows the
variation  of  shear strength with  the viscosity of pore fluids.
According to Figure 4, a pore fluid with viscosity values ranging
from  one to  three times that of distilled water makes the soil
stronger. If one uses pore fluids with a viscosity three times greater
than distilled water, the sample is weaker and when the viscosity
is greater than ten times that of distilled water, the soil becomes
much weaker than when it was compacted with distilled water.

                            Table 4
                   Results of the Strength Test
                    And Pore Fluid Viscosities
Pom
Fluid
Type
Distilled Water
25% 1-Propanol + Water
50% 1-Propanol + Water
1-Propanol
1-Butanol
16% Glyoerol + Water
28% Glyoerol + Water
Shear
Strength
C^ in KPa
5.4
10.9
26.1
45.5
36.5
1.14
0.98
Dry
Density
g/ca?
1.25
1.35
1.32
1.32
1.25
1.28
1.27
Viscosity
(cp)
at 25ฐC
0.80
1.16
1.53
2.26
2.95
158.4
272.0
                                                                                   CONTAMINATED SOIL TREATMENT    387

-------
                                                                         0.4
                            Figure 4
           Variation of Strength Ratio with Viscosity Ratio
  This behavior may be explained as follows. With the increase
in pore fluid viscosity, the lubricating nature increases and hence
the strength decreases. The peak strength was obtained from an
increase in grain interlocking. As stated in the earlier section on
compaction, with the increase in viscosity, the compaction curve
moves to the right. Therefore, for a given dry density, there will
be higher degree of interlocking of the panicles (as selected density
approaches the optimum value). This higher degree of interlocking
results in higher strength. Beyond the optimum shear strength, the
lubricating nature of the fluid dominates over the grain interlock
and hence the shear strength is reduced. Beyond this point, the
compaction curves  move to the left.

Slope of Isotropic Consolidation Line.X
  Bolt* attempted to predict the compressibility characteristics of
clays  based on  the concept of osmotic  pressure using Gouy-
Chapman diffuse double layer theory and Van't Hoffs theories
of parallel platy particles. The theory, however, was found to be
valid only for clays exhibiting very strong colloidal properties such
as montmorillonite. It has been shown  by Rosenqvist14 that the
compressibility of clay is dependent on the type as well as the con-
centration of the pore fluid.
  The factors influencing the magnitude of dielectric dispersion
Aco were  investigated in detail by Arulanandan et  al.3.  They
found that the dielectric dispersion is significantly influenced by
the type and amount of clay mineral.  The values of Aeo were in-
creased in the sequence kaolinite < illite < montmorillonite. The
compression index of these soil also increased in the same sequence.
The magnitude of dispersion decreased with an increase in per-
centage of sand in sand-clay mixtures and so does the compres-
sion index2. Olson et. al.13 showed that the compression index of
kaolinite decreased when the electrolyte concentration increased
from 0.0001N sodium chloride to l.ON sodium chloride Arulanan-
dan et al.2  showed  that Ato also decreased with increasing elec-
trolyte concentration. Based on the above data, Arulanandan et
al.3 proposed the  correlation,  shown in  Figure 5, between the
slope of the isotropic consolidation line X and the magnitude of
dielectric dispersion, Aซo.
  All the pure soils used in this investigation and a few contami-
nated soil test results are plotted on Figure 6. It appears a correc-
tion is needed for high or low pH soil samples and for soils with
pH of approximately 7 agrees well with the above correlations.

388    CONTAMINATED SOIL TREATMENT
                                                                         0.3
                                                                      o
                                                                      5
                                                                      8 02
  u.
  o
  UJ
  a.
  o
                                                                         O.I -
             • Natural Soil
             a Snow Col + 5% Montmorilldnite
             AYolo Loom
             oSnow Col + Illite
             VMorysville Red Soil
             x Illite
             *Kaoline - MP
             •••Mixed  Soils
        0      10      20     30     40     50     60    70

           MAGNITUDE OF DIELECTRIC DISPERSION,  AE0

                            Figure 5
      Correlation Between Compression Index. X and Magnitude
     of Dielectric Dispersion, Aปo (after Arulanandan et al, 1983)
                                            • Sell I

                                            • Soil II

                                            • Soil I]

                                            - fc.ii it

                                            • SOU I)

                                            A Soil 12 vltfc 251
                                             rropuol • JS1 Bป
                                           ** SoU 13 ปlt' ซซC1
                                           A Soil I) wlU Bad
                  lUttiltvKli of DUloctru Diioinlm, it

                            Figure 6
          Variation of Compression Index, X and Magnitude
         of Dielectric Dispersion, A.o for Contaminated Soils

CONCLUSIONS
  The experimental test results suggest that a completely different
approach is needed in interpretation of physical properties of con-
taminated soils. For compaction, the density and the volume frac-
tion of the  contaminant fluid should be included in data analysis.
The strength is strongly dependent on the viscosity of the pore fluid!
  For a meaningful identification and characterization of contami-
nated soils, the electrical method is found to be much superior than

-------
the current methods. The dielectric dispersion may be used as an
index for such identification and characterization. The electrical
properties should be corrected for the variations in pH value. We
recommend that further research to be performed for complete
characterization of contaminated soils.

ACKNOWLEDGEMENTS
  The study described in the preceeding pages was supported by
a research grant from the New Jersey Department of Higher Edu-
cation and the Foundation of New Jersey Institute of Technology.
This support is gratefully acknowledged. The authors are grateful
for the assistance  given by Mr. T. Mahendraratnam.

REFERENCES
 1. Anandarajha, A, "In-Situ Prediction of Stress-Strain Relationships
    of Clays using a Bounding Surface Plasticity Model and Electrical
    Methods," PhD Dissertation to the University of California, Davis,
    CA,  1982
 2. Arulanandan, K. and Smith, S. "Electrical Dispersion in Relation to
    Soil Structure, J. SMFE, ASCE, 99, SM12, Dec 1973.
 3. Arulanandan, K. Anandarajha, A. and Meegoda, N. J. "Soil Charac-
    terization for Non-Destructive In-Situ Testing," Symposium, The In-
    teraction of Non-Nuclear Munitions with Structures, Part 2, U. S. Air
    Force Academy, CO, May 1983, 69-75.
 4. Bolt, G. H. "Physico-Chemical Analysis  of the Compressibility of Pure
   Clays", Geotechnique, 6, 1956 86-93
 5. Chapman, D. L., "A contribution to the theory of Electrocapillarity,"
   Philosophical, Mag, 25, 1913 475-481.
 6. Gouy, G., "Sur la constitution de la charge electrique a la surface d'un
   electrolyte," Anniue Physique (Paris), Series 4, 9, 1910 457-468
 7. Lambe, T. W. and Martin R. T. "Composition and Engineering proper-
   ties of Soil I," H.R.B. Proc. 32, 1953 576-590
 8. Lambe,  T. W. and Martin,  R.  T.  "Composition and Engineering
   properties of Soil II", H.R.B. Proc. 33,  1954 515-532.
 9. Lambe,  T. W. and  Martin  R. T.  "Composition and Engineering
   properties of Soil III," H.R.B. Proc.  34 1955 566-582
10. Lambe,  T. W. and Martin,  R.  T.  "Composition and Engineering
   properties of Soil IV", H.R.B. Proc.  35 1956 662-677
11. Lambe,  T. W. and Martin,  R.  T.  "Composition and Engineering
   properties of Soil V," H.R.B. Proc. 36,  1957 693-702
12. Meegoda, N. J. "Prediction of In-Situ Stress State using Electrical
   Method," M .S. Thesis of the University of California, Davis, CA, 1983.
13. Olson, R. E. and Thompson, C. D., "Mechanisms Controlling Com-
   pressibility of Clays," /. ASCE, of Soil Mech. and Foundation Div.
   SM6, Nov. 1970.
14. Rosenqvist, I. Th., (1958) "Physico-Chemical properties of Soils-Soil
   Water Systems," J of SMFE, ASCE,  85, SM2, April 1958, 31-53.
15. Smith, S and Arulanandan, K. "Electrical Dispersion in Relation to
   Soil Properties," J.  GED, ASCE, 107, GTS, May 1981.
                                                                                    CONTAMINATED SOIL TREATMENT    389

-------
                   INSITU,  Vacuum-Assisted,  Steam  Stripping  of
                                      Contaminants  From  Soil

                                           Arthur E. Lord,  Jr., Ph.D.
                                        Robert M.  Koerner, Ph.D., P.  E.
                                                Vincent P. Murphy
                                         Geosynthetics Research Institute
                                                 Drexel University
                                                 Philadelphia,  PA
                                             John  E. Brugger, Ph.D.
                              Hazardous Waste Engineering  Research  Laboratory
                                    U. S. Environmental Protection  Agency
                                                     Edison, NJ
ABSTRACT
  Work is described in regard to development of an insitu, vacuum-
assisted steam stripping method to clean up contaminated soils.
Various topics are discussed including:

• The volume change of various soils exposed to steam.
• The behavior of a steam front in a thin, transparent box, from
  which steam  permeability coefficients can be estimated for a
  variety of soil types; this permeability is compared to that of
  water permeabilities in the same soil; the movement of the front
  seems to be governed by the water permeability.
• The behavior of a steam front in stratified soil
• Small scale (500 cm3 reaction flask) experiments on steam
  stripping of gasoline from two different soils; the method is
  100% efficient in sand in 3 hr and also relatively efficient (80%)
  in a 50% silt/50% sand mixture in  the same time frame.
• The preliminary performance of a unique geosynthetic soil cap
  assembly; this method appears quite promising since all of the
  steam injected below the cap was contained by the cap system;
  presumably the steam can be efficiently extracted using standard
  equipment  and methods.

INTRODUCTION
  Whenever soils are contaminated by chemical spills (such as high-
way or railroad accidents), it is important that the chemicals be
prevented from reaching the groundwater. Fortunately, in many
locations the partially saturated or vadose zone exists and acts as
temporary containment retarding the downward movement of the
pollutant. The remediation options are:

• Excavation and off-site disposal
• Excavation and on-site treatment
• Insitu treatment (via a number of possible methods, e.g., bio-
  logical, physical or chemical)

  A number of these techniques (and others) have been reviewed
in recent articles (1,2).
  The present study falls in the insitu treatment category wherein
the authors propose to have pipes inject steam into the soil beneath
the contaminated zone. Steam stripping of the chemical occurs and
when aided by a vacuum at the ground surface brings the con-
taminants to a collection point where they can be properly treated.
A unique aspect of the study is a geosynthetic cap assembly con-
sisting of a high transmissivity geotextile and a flexible membrane
liner (geomembrane). The vacuum is applied to the underside of
this liner and  the contaminated gas and/or liquid moves beneath
the liner in the geotextile to the outlet ports. A schematic diagram

390    CONTAMINATED SOIL TREATMENT
of a proposed system is given in Figure 1 for reference purposes.


                                     A*
             ซpป UmrtoU Syซtซm

                       PUkNVIEW





            - Anchor Trancfi
vtj Vacuum
./   ^.     \  ,— Gtommtirm
                                     GroundwMr TMX*
                     ELEVATION VIEW
                        Figure 1
            Schematic Diagram of Proposed Insilu.
Vacuum-Assisted Steam Stripping Field Apparatus
  The overall research project is divided into the following phases:

Phase I — Run small scale laboratory experiments to determine
    the feasibility of steam stripping of a wide range of chemicals
    for pertinent soil types and their different conditions.
Phase II — Design and perform small pilot-scale steam stripping
    and vacuum entrapment experiments on selected chemicaVsoil
    combinations.
Phase III — Perform final design of field unit. Possible super-

-------
    vision of the construction of a field deployable unit working
    closely with U.S. EPA-Edison personnel.

  Although a great deal of chemistry and chemical engineering
literature is  available  as  regards steam  stripping and  steam
distillation3'4, there is virtually no literature concerning the inter-
action of steam with soils. Thus the preliminary work has involved
relatively simple experiments to develop basic knowledge in this
area.
  The volume expansion of a variety of soil types exposed to steam
(at about 100ฐC and 1 atm pressure) in an autoclave was measured.
The behavior of a steam front (injected at about 100ฐC and 6
lb/in.2) in a  soil-filled, thin transparent  box  was determined.
From these steam front movements,  effective permeabilities for
steam movement was derived and compared with water perme-
abilities for a variety of soils. Also the effect of soil stratigraphy
on  steam front movement was demonstrated.
  Small-scale (500 cm3 reaction flask)  steam stripping experiments
are described with gasoline as the contaminant. Containment of
the stripped waste is an absolutely essential aspect of the work.
The geosynthetic containment system is of a unique design, and
some preliminary results with this containment system are given.

EXPERIMENTAL DETAILS

Changes During Steam Exposure
  Plastic containers (6-in. inner diameter) were filled with soil to
a depth of about 6-in. and were exposed in a steam-filled autoclave
(100 ฐC and 1  atm pressure) for various times up to 480 hrs. Volume
changes were determined periodically with  a simple dial  displace-
ment gage set up before and after the sample had cooled. A variety
of soil types and soil conditions were used.  The soils ranged from
stone aggregates to clays at various densities  and moisture  contents.

Steam Front  Movement
  Figure 2  shows a picture of the two-dimensional steam front
movement observational cell. (A water trap has been added since
the picture.)  The plexiglass cell is 3 ft x 2 ft x 1 in. thick.  (After
repeated runs with the steam, the plastic bowed-out and the thick-
ness increased.) Steam, from the laboratory lines at approximate-
ly 6 lb./in.2 100 ฐC was injected into  the center of the soil-filled
cell and measurements were made of the expanding wet front versus
time.  Again  a variety  of typical  soils were  used. The cell was
instrumented with  temperature  sensors on a 6-in square array.
  The same cell was used to observe the effect of soil stratigraphy
on  steam front movement. This measurement was accomplished
by placing layers of different materials in the cell. In all these cell
observations, the soils  were placed in 4-in. lifts and compacted.
Small-Scale Steam Stripping of Gasoline
  Figure 3 is a schematic diagram of the small scale device used
to investigate the efficiency of steam stripping of gasoline from
soil. Steam generated from a home-style pressure cooker is fed into
the bottom of a 500 cm3 reaction flask containing well mixed soil
and gasoline. The steam/gasoline vapor mixture which is stripped
from the soil enters a distillation  column  and the condensed
material is collected in another flask. In future experiments, vacuum
will be applied at the end of the distillation column. The amount
of gasoline stripped  was determined by allowing the condensed
water/gasoline liquid  mixture to settle  for at  least 3  hr in a
graduated cylinder and then determining the amount of gasoline
volumetrically.
                                         cooling
                                         water
                                                    condensott
      HEATED
  STEAM  GENERATOR
REACTION
 FLASK
DISTILLER/COLLECTOR
                          Figure 3
      Schematic Diagram of Small-Scale Steam Stripping Setup
Geosynthetic Containment System
  Figure 4  is a schematic diagram  of the  device used in the
preliminary  attempts at steam capture in soil. This cap assembly
consisted of an HOPE geomembrane formed into an open-ended
"box" of dimensions approximately 1 ft x 1 ft x 1 ft. A thick poly-
propylene needle-punched geotextile was placed inside the geomem-
brane cap. A vacuum connection was made at  the top and the
assemble was imbedded into dry sand to an anchor trench depth
of about 6-in. A cold trap of very crude design was placed between
the cap and the vacuum pump. Laboratory steam was injected into
the bottom of the sand box 6 in. below the bottom of the anchored
cap assembly.
   efOSYHTHCTIC
         CAP
                                                                                    CAP
                                                                        lUBfDDfD IN SAND
                                                                                                COLD TRAP
                                                                                                              VACUUM PUMP
                          Figure 2
           Photograph of Thin Plexiglass Steam Front
                 Observation Cell (without soil)
                                                                                             Figure 4
                                                                       Schematic Diagram of Setup to Observe Steam Containment
                                                                                    In Soil by Geosynthetic Cap
RESULTS
  Soil Volume Changes During Stream Exposure
  Figure 5 shows volume change results for loosely and densely
compacted well graded sand as a function of time in the steam auto-
clave. All of the  data  will not be  presented (in the manner
of Figure 5) here for all the soil types due to their considerable
volume. However, Table I indicates the general results. A few com-
ments are in order. Little water ponding was observed on the top
of any of the samples and cracking of the sample surface was not
observed to any great extent. The lack of any large change in the
silt and clay soils could be due to the relatively short exposure times
(with respect to  pertinent water diffusion times).
                                                                               CONTAMINATED SOIL TREATMENT     391

-------
   + 20
1
    -10
   -10
                     Otntfly-Compocttd  Granular  Soil
10             100

   TlUf I Houri I
                                                  [000
              Loostly - Compocrtd
                Granular  Soil
                             Figure 5
          Volume Change of Granular Soil Exposed to Steam
                             Table 1
             Results of SoU Volume Change Experiments


  Granular Sollป (grain ปUซ:  dl<  > O.Ot  ~>

       - loose; Large change  dla . >  O.OOJ mm)

       - low Mซtซr content ;  Sซ*ll chซnqซ  (lnec*aaซl
       - hlfh water contanc; Nodarat* chan^*  (lncrซaปซ)
•  CUy Sit* Solli (fraln •!ป:  dlซ. < 0.002 ซ•)


       - lew wซtซr concent; Seull  change (increซ*e)
       - high weter content; Moderate chanqe

  Coevent* on Cvperleent* :

       - sLmatl no water pondlno.
       - cracklnq very alnlatal - If present, confined to iMedlate surface

otei :
  teall chanoe     1 I ซ <  SI voluiM change
  Moderate  change  5 S ซ <  10% volume change
  Laroe change     2 lot voluaป change
                              Figure 6
                      Steam/Wet Front in Sand
               (30 seconds after start of steam injection)

392     CONTAMINATED SOIL TREATMENT
                                                   Steam Front Movement
                                                     Figures 6, 7 and 8 show photographs of various stages of tne
                                                   steam front movement in sand. The fronts looked similar in tl
                                                   silty sand soil mixtures, except that the circle moved out slower
                                                   as the silt percentage increased. At 100% silt there was no uniform
                                                   circular movement, but rather a very localized movement up a
                                                   steam-produced crack (see Figure 9 for this type behavior in 100%
                                                   silt.)
                                                                                  Figure 7
                                                                         Steam/Wet From in Sand
                                                                  (120 seconds after sun of steam injection)
                                                                                                     Figure 8
                                                                                             Steam/Wet From in Sand
                                                                                       (500 seconds after start of steam injection)
                                                                               Figure 9
                                                              Steam/Wet Front Movement in 1004k Silt Soil

-------
  If the steam front (or more correctly—the "wet" front) moves
out in a reasonably uniform circular fashion, a crude estimate can
be made of the steam front permeability. Figure 10 shows steady
state Darcy-like lateral flow between an inner cylindrical space of
radius a and pressure Pj and an outer cylindrical rim of radius b
and pressure P2. The pressure difference P,l—P2 is converted to
an equivalent hydrostatic head Ah. The thickness of the cylinder
is d and a total quantity of material q (volume/time) flows through
the annular region.  (This phenomenon  is indicated in Figure 10
by  the outflow.)  In this case  the permeability is given  by the
standard result
                                                                                        Table 2
                                                                              Steam and Water Permeabilities
k =
  _ q In (b/a)
      2,dDh
                                                  (1)
  It is assumed here that this steady state result can be applied
approximately to a slow moving transient front. Thus the approx-
imate permeability of the steam front, Ksteam, is given by
_ q In (r/a)
   20 d Dh
                                                        (2)
  It is assumed here that this steady state result can be applied
approximately to a slow moving transient front. Thus the approxi-
mate permeability of the steam front, ksteam, is given by

q = the measured volume flow rate (determined by assuming 100%
saturation and monitoring the dimensional change of the wet front
with time)
r = the particular value of the radius of the front
a = the effective radius of the input steam  "bulb"

  It is assumed here that a = 1-in. and that the entire pressure head
Ah is developed only across  the wet front.  The results  of  this
approach on a variety of soils of widely varying water permeabilities
are shown in Table II. Table II shows a comparison of the ksteam
with the standard liquid water  permeabilities kwater. It appears,
that within the accuracy of our measurements (more accurate meas-
urements of ksteamsteam  are planned) the two k-values are about
the same. That is, the movement of the steam wet front is governed
essentially by the water permeability and not the much larger dry
steam permeability5. As  noted earlier, in the 100% silt, the steam
wet front  did not move at all uniformly, so no result for  ksteam
can be given.
              OUTFLOW
                           Figure 10
      Geometric Model Used in Determining Steam Permeability
Soil
100 ป Sand
75% sand/25* silt
SOt sand/50% lilt
25% sand/25% silt
100% lilt

3 x 10~3
< X 10"'
1 x 10"4
1 x 10"4


B
1
7
3
3.
,.,. (cm/secl
x 10"3
.4 x 10"4
.1 K 10"5
.5 K 10"5
.4 x 10"6
k3team/kuat(,r
0.37
2.8
1.3
2.9

  An indication that dry steam does not move much ahead of the
wet front is given by the temperature profile of Figure 11. Here
we have plotted the temperature at a particular thermocouple as
a function of the distance from the wet front to the thermocouple.
The very sharp front indicates that dry steam (and.the associated
temperature increase) does not seem to precede the observed wet
front.
  As  an aside, we examined the effect  of soil stratigraphy.
Figure 12 shows the steam wet front in.sand as it encounters a thin
(4-in.) layer of silty sand (50% silt/50% sand). As is shown the
steam wet front has a great deal  of difficulty in penetrating the
lower permeability silty sand. More work is planned in this regard
to investigate realistic situations liable to be encountered in the field.
                                                                  100 -
                                                                  90
                                                                  80
                                                                  70
                                                                  So
                                                             .ฐ   50

                                                              LJ
                                                              CC.

                                                              I   4ฐ
                                                              CC
                                                              LJ
                                                              2:   30
                                                                   UJ
                                                                       20
                                                                       10
                                                                         -1.5   -1.0   -0.5     0    0.5    1.0     1.5

                                                                       DISTANCE  OF  FRONT  FROM  THERMOCOUPLE (inches)

                                                                                           Figure 11
                                                                        Temperature Profile Associated with Steam/Wet Front
                                                            Small-Scale Steam Stripping of Gasoline
                                                              Figure 13 shows the results of steam stripping of unleaded gaso-
                                                            line from two soil types of low moisture content. The initial ratio
                                                            of gasoline to soil was about 0.15 by volume in both cases and
                                                            thus the pore spaces of the initially dry soil were not filled with
                                                            gasoline. For the sand, within the limits of our  somewhat crude
                                                            analytical technique, we found that all the gasoline was stripped
                                                            in 3 hr. In the 50% silt/50% sand mixture, over 80% is stripped
                                                            in less than 3 hr. Drier steam and/or vacuum distillation might
                                                                               CONTAMINATED SOIL TREATMENT     393

-------
have increased the yield in the silt/sand case. This variable will
be the subject of future work.
                           Figure 12
               Behavior of Steam/Wet Front in Sand
             When it Encounters a Layer of Silty Sand
      100
 2
 Uf
 K
 UJ
 Z
               30
                      60
                                                        210
                           TIME (mii
                         Figure 13
           The Percentage of Gasoline Steam Stripped
           From Two Soil Types as a Function of Time
  The authors know of no other direct laboratory studies of steam
stripping of chemical  from soils. The analytical Nielsen-Kryger
steam distillation procedure to analyze residual chemical contents
of soil is closely related6. Also a large steam/hot air soil decon-
tamination system has been developed and is being tested1.

Geosynthetic Containment System
  Figure 4 shows a schematic diagram and Figure 14 a photograph
of the equipment to test  the steam containment  capability of the
geosynthetic containment cap (Figure 14 does not how the water
trap in place). Laboratory steam was admitted into the bottom of
the sand box directly  below the geosynthetic cap assembly. The
steam ran for about 30 min and no steam escaped around the edges
of the  cap. Unfortunately our trap (simply a flask imbedded in
ice water) was not  very  efficient and much steam was  observed
exiting the vacuum pump; a vacuum of 10-in. Hg. could be main-
tained  at the vacuum pump end of the system. Although a great
deal of development work  needs to be  done, it was shown that
anchoring of the geosynthetic cap by 6-in. imbedding is sufficient
to contain  the rising steam.
                                                                                            Figure 14
                                                                        Photograph of the Setup to Determine the Steam Capture
                                                                         Ability of • Geotynthetk Cap in Soil (Trap not Shown)
CONCLUDING REMARKS
  Steam stripping of contaminants from soils seams to be a viable
insitu cleanup technique in principle.
  To apply the method efficiently the following must be known:
• Permeability of steam/water in various soUs
• Gross behavior of  steam/water fronts in various soils
• Efficiency of steam stripping of various contaminants in a variety
of soils

A viable geosynthetic containment cap/vacuum assembly must be
developed. From the work described in this paper it appears to
the authors that this information can be developed, and  to a
reasonable probability a field system can be developed.
  In view of the previous remarks in this Section the following
future work is planned:

• More accurate measurements  of  the steam  permeability are
  needed. To this end a cylindrical geometry permeability setup
  of known  fixed  input and output dimensions has been  con-
  structed where the  time for the temperature front to traverse a
  known radial distance can be determined accurately. Also the
  input and output steam volume flow rates can be determined
  and hence  a reliable permeability  can be determined by more
  conventional  means.
• Additional  small-scale steam stripping measurements will be con-
  ducted using gasoline (and other contaminants) in a wide variety
  of soil. Vacuum-assisted steam stripping will  be used.
• Work on the geosynthetic cap assembly will proceed in earnest.
  The development of a steam condenser is very important here.
• It would be of definite interest to use higher temperature/higher
  pressure steam (i.e., drier steam) in all the described aspects of
  the work. This experimental work may prove to be quite difficult
  however at this early stage of our work.

ACKNOWLEDGEMENTS
  The Drexel  authors  wish  to  thank the Hazardous Waste
Engineering Laboratory of the U. S. EPA of Edison, New Jersey
for their financial support of this work through Cooperative Agree-
ment No. CR-813002-01.

REFERENCES
1. Kovalic, J.  M. and Klucsik, J. F., "Loathing for Landfills Sets Staft
   for Innovative Hazardous Waste Treatment Technology " Hazard MA
   & Waste Manag. 5.  1987, 8, 17-18.
2. Cheremisinoff, P.  N., "Update: Hazardous Waste Treatment " Polhtl.
   Eng. 19. Feb. 1987,42-49.
394     CONTAMINATED SOIL TREATMENT

-------
3.  Hwang, S. T. and Fahrenthold, P., "Treatability of Organic Priority          Res Res, 19, 1983, 931-937.
   JJfSg8 g Seam ^PP^'" A.I.Ch.E. Symposium Series, 76, No.       g  DeV( R  (
-------
                  Innovative  In  Situ  Soil  Decontamination  System

                                              Richard  Van  Tassel, Ph.D
                                                   NUS Corporation
                                               Pittsburgh, Pennsylvania
                                                 Mr.  Philip  N.  LaMori
                                               Toxic  Treatment Limited
                                                 San Mateo,  California
ABSTRACT
  An innovative system has been developed to treat soils and wastes
on-site and in-situ. This system incorporates procedures and equip-
ment that are appropriate to  reduce the concentration and/or
mobility of contaminants. The processes employed by this equip-
ment consist of oxidation, pH adjustment, steam/air stripping, off-
gas scrubbing and stabilization/fixation. These processes are appro-
priate for treating soils and wastes to  reduce the toxicity of the
treated materials. Special equipment has been developed to per-
form the treatment operations in-situ and on-site.
  This in situ treatment system recently was employed to reduce
the concentration of hydrocarbon contaminants from soils at a site
where fuels had leaked into the  underlying  soils. The results of
analytical tests of soil samples obtained before and after treatment
indicated that a greater than 90% reduction in hydrocarbon con-
tent was achieved. Initial hydrocarbon content in soils ranged from
1,000 to 40,000 ppm. The reduction in hydrocarbon content was
sufficient to permit re-development of the site for warehousing,
freight transfer and office building land use.
  The in situ  treatment of contaminated soils and wastes offers
many advantages that are appropriate for site remediation. These
include less excavation, the destruction and/or immobilization of
contaminants, no transport of contaminated material and no long-
term storage/management requirements. These advantages should
result in a reduced cost for site cleanup. The capabilities and oper-
ations of this in situ treatment system, as demonstrated at its first
site remediation project, are discussed  in this article.

INTRODUCTION
  The capability to remove hydrocarbon materials from soils has
been provided by assembling appropriate processes  into an inte-
grated, soil detoxification system.  The system provides for the
detoxification of the soils in situ,  and the necessary operations are
performed on-site. Accordingly,  this system provides an alterna-
tive to other remedial technologies which require the excavation
of the contaminated soils and further processing or removal.
  This integrated system has been  designed and developed by ATW
and Calweld of Sante Fe Springs, California. It is being marketed
in the U.S.A. by Toxic Treatment  Limited of San Mateo, Cali-
fornia. The use of this equipment to remove hydrocarbons from
underlying soils, contaminated by leaded and/or spilled fuels at
a truck servicing area, was demonstrated at a site near Long Beach
California.  The hydrocarbon  contaminants were reduced suf-
ficiently to permit the on-site development of a new freight-transfer
facility.
  This article describes the processes and equipment utilized on
this demonstration project. The capabilities of this system are
presented. In addition, the potential for treating hydrocarbon- con-
taminated soils using this on-site and in situ equipment are iden-
tified.

IN SITU SOIL  DECONTAMINATION EQUIPMENT
  The soil detoxification system  employs in  situ operations to
remove hydrocarbon contaminants and/or to fix inorganic con-
taminants in a stabilized soil mass. The equipment used to perform
these operations is described in this section. This equipment and
its relative positioning on-site are  identified in Figure 1,  "Plan
Layout of Equipment."

Process Tower
  The key unit for the in  situ treatment operations is the "Process
Tower." This unit provides the penetration into the underlying soils
and conveys/mixes reagents into the soil mass being treated. This
unit is shown in Figure 2.
  The process tower is comprised of a drill tower carrying two
hollow "kellys" mounted atop an insulated steel box called the
"shroud." The  shroud rests on the ground surface and forms a
seal to prevent the escape of gasses during the operations. Two
counter-rotating turn tables are mounted  atop the shroud, and
hydraulic motors drive two turn tables to rotate the kellys. Auger
bits, mounted on the bottom of the kellys, loosen and mix the soils
as the kellys rotate and advance into the ground. Two hydraulic
motors power a  pull  down drove to raise/lower the kellys and po-
sition the auger bits at the desired depths.
  The kellys are about 30 ft long and transmit the torque from
the turn tables to the auger bits. Keys on the outside of the kellys
are engaged by the turn tables to receive the torque as the kellys
traverse up and down. Inside the hollow kellys there are dedicated
conduits that convey wet reagents to liquid jets located on one blade
of the auger bit. Other conduits convey heated air and dry rea-
gents to a port on the opposite side of the augers for expulsion
into the soil mass.
  The augers are two bladed, providing a bit diameter of 4-1/2
ft. The two auger blades are arranged to overlap and rotate in oppo-
site directions.  Collisions of the blades are  prevented by syn-
chronized rotation of the  turn tables and a separating yoke located
approximately 2 ft above  the blades. The overlapping augers allow
the treatment of a 30-ft'  area of soil through the 27-ft.deep range
of penetration. This arrangement allows the treatment of  1.1 yd'
of soil for each foot of  penetration.

Carrier Tractor
   Mobility is provided to the "Process  Tower"  by the "carrier
3%    CONTAMINATED SOIL TREATMENT

-------
                                                                  / WflCHEMICAL
r                                                                      PUMP TRUCK
                                                                     F=n'
                                       CHEMICAL STORAGE
                                                      OO
                                                      OO
                                     iQ.QQogg_	88
                                	 _^__ 	TOTAILt WATCH ^_        __ 	J
                                                                                              DRY
                                                                                            CHEMICAL
                                                                                             STORAGE
                                                                                             TRAILER
                                             Figure 1
                                       Layout of Equipment
           Figure 2
 Carrier Tractor, Process Tower,
Shroud, Operations Control Booth
                                                   tractor" which is also shown in Figure 2. The carrier tractor is a
                                                   modified "Caterpillar" pipeline tractor complete with adjustable
                                                   counterweight to balance the process tower.
                                                     The process tower is connected to the carrier tractor by articu-
                                                   lated arms. Hydraulic rams in these arms afford some control of
                                                   the  tower  by the carrier tractor operator.  The operator can
                                                   raise/lower, thrust in/out, rotate and plumb the process tower to
                                                   place the shroud at the proper location and orientation.
                                                     The process control booth is mounted on the front end of the
                                                   carrier tractor. A hydraulic power pack, which supplies hydraulic
                                                   pressure and fluid flow to power the hydraulic motors on the
                                                   process tower, is mounted on the rear of the carrier tractor. These
                                                   units also are shown in Figure 2.
        Figure 3
Off-Gas Scrubbing System
                                                                CONTAMINATED SOIL TREATMENT    397

-------
 Process Control Booth
   The  process control booth  houses control and  monitoring
 instrumentation so the operations director can control the opera-
 tion of the equipment and evaluate the soil treatment procedures.
 Control valves are available for the operator to control the flow
 of steam and hot air for stripping the soil block. Controls also arc
 provided to control the dry chemical feeding equipment from the
 control booth.
   Process monitoring equipment is provided in the control booth
 so the operations director can evaluate the operation of the equip-
 ment and the treatment of the soil block. Pressure and  tempera-
 ture  gauges are  provided  to  monitor the  operations  of the
 equipment. Switches are provided to control the position of valves
 of the various equipment.
   Monitoring equipment is provided to monitor the treatment of
 the  soil block and removal  of hydrocarbons from  the off-gas
 stream. A flame ionization detector and temperature gauges were
 provided to monitor the concentration of hydrocarbons and tem-
 perature in the off-gas and return air  (after the activated carbon
 Hlter) streams. In addition, temperature.  pH and redox sensors
 were mounted on a probe that monitored soil conditions about
 18 in. above the auger bits. A strip chart recorder plotted the con-
 centration in the off-gas and return air streams. Manual record-
 ing of other operations and treatment data were entered on a log
 for the soil block  being treated.

 Off-Gas Scrubbing System
   The off-gas scrubbing system is a trailer mounted unit that moves
 with the mobile Process Tower. The contaminants are removed
 from the off-gas by means of condensation, density separation and
 activated carbon filtration. The trailer-mounted off-gas scrubbing
 system is shown in Figure 3.
   Two refrigeration coils  are provided to cool the off-gases and
 to condense the vapors in the gas stream. After passing through
 the first refrigeration coil, the off-gases pass  through two series
 connected cyclone separators. Condensate droplets are separated
 from the gas stream and are collected in two storage tanks located
 under the separators.
   The off-gases then flow through a second refrigeration coil before
 passing through an  activated carbon filter.  The Tiler has two
 separate bins, one  containing granulated activated carbon (GAC)
 and the other either GAC or shredded coconut hulls. The off-gases
 exit  the activated  carbon  filter and then pass to the return air
 handling system.

Return Air System
  The scrubber off-gases are returned  to the soil block for more
stripping operations by means of the return air system. This system
is mounted on a trailer and is towed behind the carrier tractor.
 Figure 4 shows the mobile return air system which consists of an
 air compressor and two air heaters. The dry chemical feed system
 also is mounted on this trailer.
   The return air stream is first conveyed to a piston type air com-
 pressor to raise the pressure for reinjection. The air compressor
 is powered by an air cooled diesel engine of approximately 200 hp.
 The compressor handles up to about 500 ft'/min and increases the
 return air pressure to about 350 Ib/in.2 and increases the return
 air pressure side of the air compressor in two parallel streams as
 it is conveyed to the two kellys for more in situ  stripping.
   Each return air stream passes through a spherical heat exchanger
where the temperature of the returning air is increased to about
315 ฐF to enhance the in situ stripping. The two air heaters use heat
exchangers to recover heat from the regrigeration system and elec-
trical resistance heating elements. The heated air either returns
directly to the process tower for more in situ stripping or to the
dry chemical feed  system, if the dry reagent option is selected.

Dry Chemical Feed System
   Two parallel dry chemical feed systems are provided, one for
each return air stream. These units are shown in Figure 4. The dry
chemical feed system consists of two storage bins, two screw fedders
and two air/powder mixers.
   Parallel conduit and valve systems  are provided to the return
air stream to bypass the dry feeders if the air stripping cycle U
ongoing. If  the soil stabilization operation is being performed,
appropriate valves are open/closed to pass the returning air stream
through the air/powder mixers to entrain and convey dry material
to the process tower to stabilize the soil block.

Portable Boiler
   Steam required for the in situ stripping operations is provided
by a trailer-mounted boiler. This unit  is shown by Figure 5. This
boiler moves parallel with the process tower during operations.
   The boiler supplies superheated steam at the rate of 2500 Ib/hr.
The steam is heated to 350 T and at a  pressure of 360 Ib/in.2 The
superheated  steam is conveyed to the separate kellys where it is
transported through the kellys in conduits dedicated to convey wet
materials. At the auger bits, the steam passes through jets on one
blade of the auger and is injected into the soil block being treated
by in situ stripping.
                           Figure 4
            Return Air and Dry Chemical Feed Syitems
                                                                                             Figure 5
                                                                                          Portable Boiler
Liquid Chemical Feed System
  Liquid chemicals were supplied to the drill tower for injection
into the soil mass by means of a liquid chemical supply truck. To
fulfill site requirements, a liquid chemical mixing/storage tank and
a high  pressure pumping system were  leased from a  service
company.
398     CONTAMINATED SOIL TREATMENT

-------
  The liquid chemical supply truck is positioned at one location
on- site and the flexible conduit supplies the mobile equipment as
it traverses the site. This system could be mobile if necessary, but
required chemical stores make it advantageous to allow this equip-
ment to remain stationary.

Dry Chemical Storage
  Dry chemical storage was provided on-site from trailer-mounted
storage bins obtained from the service company. The trailer con-
tains 3 storage bins, a cyclone for suspending the dry chemicals
in a stream of air and a diesel engine driven air compressor. The
weight of the cement stored in this unit  required that the trailer
be supported on a frame placed on  the ground surface, thereby
unloading the tires. Due to this requirement, this stationary trailer
is located where it can remain in one place.

TREATMENT PROCESSES
   A summary of the unit processes which the soil detoxification
equipment can perform is described in this section. These processes
are presented in the sequence normally  employed in the field.

Chemical Oxidation
  Chemical oxidation of petroleum hydrocarbons was performed
in situ using the soil detoxification equipment. At the demonstra-
tion project, a solution of potassium permanganate, sulfuric acid
and water was injected and mixed into the hydrocarbon contami-
nated soils. The oxidation of the hydrocarbon contaminants was
enhanced by the addition of heated air and the mixing action of
the augers.
   The oxidation of hydrocarbons using  a solution of potassium
permanganate is based  on  bench scale tests  performed  by
ATW/Calweld and experience gained by the chemical industry.
A supplier of potassium permanganate  has reported success in
oxidizing pollutants in aqueous and sludge streams. For the demon-
stration project site, ATW/Calweld performed bench-scale tests
using site soils and the fuel contaminants to determine the proce-
dures and formula for the oxidizing solution  and evaluated the
results by laboratory tests on the soils.
   Other oxidizers could be employed to reduce the concentrations
of hydrocarbons. ATW/Calweld also performed bench-scale tests
using hydrogen peroxide and chlorine dioxide. These reagents were
not employed on-site. It is stressed that bench-scale tests must be
performed according to a sound testing program to:

   Determine the quantities of reagents
   Identify reaction products
   Evaluate the cleanup results
   Understand the reaction kinetics
   Reduce the possibility of undesirable  conditions

Steam and Hot Air Stripping
   The volatilization of hydrocarbons  using hot air is a known
process. In addition, the use of superheated steam in the stripping
process is known to enhance the volatilization and stripping of
hydrocarbons from adsorbing media. The passing of heated  air
over hydrocarbons will increase the rate of volatilization because
of the greater quantity of  air passing over the materials and the
higher temperature which  increased the  rate of vaporization.
  The superheated steam enhances the removal of hydrocarbons
by imparting more heat energy into the soil mass in a short time
period and by the steam will displacing hydrocarbon molecules that
might be adsorbed on soil  particles. Enhanced recover of hydro-
carbons from soil, water, rock and adsorbing material using steam
is a widely used process employed  by petroleum and chemical
industries.
  At the  demonstration project, steam and hot air stripping was
performed in situ by forcing air and steam through the auger bits
and into the soil mass. These streams were conveyed to the  point
of injection through separate conduits in the hollow kellys. The
superheated steam was injected into the soil mass by jets mounted
on one blade of the auger. Superheated steam, 350 ฐF and 360
lb/in.2 was injected at the rate of 2500 Ib/hr.
   Heated air was blown into the soil mass through a port on the
auger located on the kellys opposite the steam jets. The air was
heated to 315ฐF and injected at rates of 325 to 750 ftVmin.  In
the final configuration of the air delivery system, using the piston
type air compressor, the air flow was at 325 ftVmin.
   The injected hot air, steam and volatilized hydrocarbons,  at
elevated pressures in the treatment zone, are captured at the surface
of the disturbed soil column by means of reduced pressures in the
"shroud." The shroud is an insulated steel box with dimensions
of about 7 ft wide, 10 ft long and 5 ft high. A flexible conduit
connects the shroud to the suction side of a blower which creates
the reduced pressure within the shroud. The high pressure side of
the blower forces the off-gases to  the on-site scrubbing unit.
Off-Gas Scrubbing System
  The off gases are passed through a multi-staged scrubbing system
to remove the vapor phases of contaminants. The scrubbing system
consists of series mounted cooling/condensing coils, cyclone sepa-
rators and an activated carbon  filter.
  The off-gases are coiled first in a cooling coil to condense the
organic vapors and steam in the gas  stream. These cooler gases
then enter two series connected cyclone separators to remove liquid
droplets from the gas stream. Liquids captured by these separa-
tors are collected in two polyethylene tanks of about 250 gal total
capacity.
  The gas stream then flows through a second coiling coil to further
condense vapors and cool the gases before entering the activated
charcoal filter bin. Condensates from this cooling unit are collected
and placed into a container for eventual disposal.
  The activated charcoal filter has two compartments within the
bin and is about 4 ft by 2.5 ft long. A diffuser and diaphragm baffle
increases the flow path and the potential for contact and removal
of hydrocarbons from the gas stream. Initially, granulated activated
carbon was used in the first compartments and coconut hulls were
used in the second compartment. Later in the site operations, granu-
lated activated carbon was used in both compartments, and 1000
Ib of carbon  was required to charge  the filer.

Return Air System
  After the off-gases passed through the scrubbing system, the
gases were recycled to the soil block to perform more stripping.
Before being  reinjected into the soil mass, the gas pressures and
temperatures  were increased to enhance the volatilization rate of
the hydrocarbon contaminants. The return air system was mounted
on a trailer that was pulled behind the carrier tractor. This trailer
and equipment are shown in Figure 5.
  The gases exiting from the activated carbon filter were conveyed
by a 6-in. diameter hose to a manifold feeding the suction side of
a two cylinder, piston type air compressor. The pressures of the
off- gases were increased to about  360 lb/in.2 at rates up to 500
ftVmin.
  The gas flow then was conveyed to two spherical heat exchangers
installed in parallel. The air heaters contained heat exchangers  to
recover heat  from  the  refrigeration system and the electrical
resistance heating elements. The returning gases were heated  to
315ฐF and transported in  separate  conduits to the two kellys  or
two dry chemical feeders mounted on the  trailer.  Thus, the off-
gases were recycled to the soil mass undergoing treatment, com-
pleting the closed loop circuit.

Steam  Supply System
   Superheated steam is supplied to the in situ treatment equipment
by a trailer mounted, mobile boiler. The boiler supplies superheated
steam at 350ฐF  and 360 lb/in.2 at the rate of  2500 Ib/hr. The
capacity of this portable boiler was reported at 3 million BTU/hr.
An on-board water system treats potable water to boiler feed water
quality requirements to control corrosion and/or scaling of the
boiler. The boiler is fired with  fuel oil from an on-board tank.
                                                                               CONTAMINATED SOIL TREATMENT     399

-------
   Superheated steam is conveyed to the kellys in separate conduits.
The steam is injected into the soil mass through liquid jets mounted
on one blade of each auger. The injected steam assists in displacing
hydrocarbon molecules adsorbed to the soil and also increases the
volatilization rate of the hydrocarbon materials.

Soil Stabilization
   After the stripping operations are completed, a stabilization
operation can be performed on the soil block. The in situ stabili-
zation operation can be performed to reduce the mobility of in-
organic substances by chemically complexing  these substances
and/or by reducing the permeability of the soil mass. The soil block
normally contains more moisture than normal due to the liquids
added and the condensation of steam during the stripping opera-
tions. In addition, the soil block has been loosened because of the
agitation action of the augers during the in situ treatment opera-
tions. The loosened and moist soil block can be stabilized by the
addition of cement, lime, fly ash or bentonite into the soil block.
   Some soil blocks were stabilized by the addition of dry cement
which hydrates water in the soil blocks. This dry material takes
up excess water in the soil and yields a firmer soil mass in the stabi-
lized soil blocks. Dry cement  is metered into the hot air return
conduits by  two screw feeders, drawing cement from two storage
bins. Hot air conveys  the cement through the kellys and discharges
it from the port in the kelly at the auger bits. The rotating augers
mix the cement into the soil mass as the augers traverse  through
the soil block  from top to bottom and return.
   The operations director monitors the feed rate of cement and
continues the injection, mixing and traversing operations  until the
                                          prescribed amount of cement has been placed in the soil mass. A
                                          typical amount of cement added to a 16:5 yd3 soil block was 629
                                          lb., distributed through the 15.3-ft depth of penetration. This was
                                          accomplished in about  12 min. Some additional time probably was
                                          required to ready the equipment for this operation which can be
                                          controlled from the control room.

                                          SOIL DECONTAMINATION OPERATIONS
                                            A brief description of the sequence of operations to treat the
                                          soil blocks is presented in this section. The time required to per-
                                          form these operations also is identified.

                                          Layout of the Soil Treatment Blocks
                                            The area to be treated by the in situ treatment equipment a laid
                                          out in a series of overlapping rectangles (3.25  ft wide  x 7.25 ft
                                          long) The overlapping rectangles are laid out in a linear arrange-
                                          ment to treat the  contaminated  soil zones  and to facilitate
                                          movement of the equipment. A typical layout  of a site is shown
                                          in Figure 6.

                                          Position Equipment
                                            The first operation required is the positioning of the equipment
                                          to perform the treatment of a soil block. In preparation for the
                                          move, the carrier tractor lifts the drill tower and shroud off the
                                          ground and adjusts the counter weight to allow the movement. The
                                          carrier tractor pulls the trailer which holds the  return air reheat-
                                          ing  units and the dry chemical feed system. The carrier tractor
                                          advances approximately 7 ft along the line of soil blocks to the
                                          next soil block to be treated.
"   r  _r    r   r   r
   rl   iM- *vr
                                                                          \~   i
                      -t   \"   ifj   \tt   itj   \f'   \"  ' V'   Yr   K^T
                         Y*—r—r—r—r*—r'  1  —r"—r*—T
                       pt    pj    jปj   L,l—i^-L/l  'pi.  pj.   pi—fi-
                        ————H*—r*—
                                                                 <—m—'  J
         —T—r—r—H—i   —i     i  —'\ —r—i  —r—r*—\^~
         J—-T—ป7—  +t    —ป—*—^—'   —'     '     '      '     I
                 r   r              r
                                   rrrr-i
                                 3	pcป	pry	rjeap	1ปป
                                 t/-*   i^>   i^y '  r^^ป  ' u
                                                       p3   L/.J—LA
                                                        —'      '
                                                                      r  T  .r  .JT"!
                                                                [/	V	p	n	ir	rr;	rr,	pf

                                                             7    IปV  T"J  Hf'7'    1"   K11   I'" T'**'  Hป1"-'
                                                                            I      '    J
                                                                       A   IT *   177    I^F


                                                                        r   r jr~r
                                                                      r   r   r     r
                                                                     -"-ป—'     '  i    '    .
                 SOIL BUXK WIIM SOIL TESt RESULTS
                 FURNISHCO TO NUS CORPORATION
                                               ~!0 FT PฃfP
                                             8UC HAP BY CALOWUA NC


                                          o      10'     to'
                                                                      SCALE
                                                           Figure 6
                                                      Layout of Soil Blocks
400    CONTAMINATED SOIL TREATMENT

-------
  The interconnected units which must move with the carrier
tractor are the boiler and off-gas scrubbing system and return air
heating trailer. The tractor pulling the trailer-mounted boiler and
scrubbing system advances parallel with a carrier tractor. These
two prime movers, the carrier tractor and boiler/scrubber tractor,
advance together at the direction of the operations director.
  After the moving units have advanced to the next soil block,
the operator of the carrier tractor positions the shroud  on the
ground and plumbs the drill tower at the direction of the opera-
tions director. The boiler steam feed valve, closed for safety during
the move, is reopened and the soil treatment system is ready to
operate. About 2 min are required to advance one soil block,
position the equipment and be ready to start treatment operations.

Initial Intrusion
  The initial intrusion of the soil block consists of the first penetra-
tion of the soil block with the augers to the prescribed depth. During
this penetration of the undisturbed soils, the downward advance-
ment of the augers is not as fast as drilling in a previously drilled,
loosened soil block.
  Hot air and steam normally are injected  through the augers
during all drilling to keep these conduits open, the reduced pres-
sure in the shroud collects  the  off-gases from  the soil block.
Monitoring of the hydrocarbon content in the off-gases provides
the operations director with an indication of the hydrocarbon con-
tent  in the soil block.
  The initial intrusion then  serves two functions: (1) the augers
advance through and loosen  the soils to permit easier penetration
in subsequent operations and (2) it allows an initial indication of
the hydrocarbon content in the soils to plan  further treatment
operations. The initial intrusion can be performed at the rate of
about 4 ft/min.

Ready Chemical Supply
  After the initial penetration, the operations  director  evaluates
the off-gas concentrations, the reported concentrations of hydro-
carbons in the area of the soil block,  and determines the amount
of chemicals to be injected at various depths of the soil block. The
current equipment feeds chemicals at one rate, approximately 50
gal/min.  Therefore, the operations  director prescribes  auger
advancement rates at various depths in the soil block to vary the
amount of liquid chemicals per unit volume of soil. These evalua-
tions/calculations may be very rapid if the soil block is similar to
previous  blocks and requires very little  time.  A  significantly
different hydrocarbon content may require 2 min. for the opera-
tions director to determine the chemical injection specifications.
  The communication of injection requirements to the drill oper-
ator and the  chemical feed  operator by the operations director
requires about 1 min. Readying the equipment for the chemical
injection requires the operator to switch only a few controls and
does not require significant  additional time.
  Therefore, the time required for this operation is about 3 min.
Computer-aided controls could make this operation nearly auto-
matic and only require a confirm signal from the operator requir-
ing about 1 min for  checking the chemical feed specification.

Inject Liquid Chemicals
  During this operation, liquid chemicals are injected into the con-
taminated soil zone according to the specifications determined by
the operations director.  The current  equipment supplies liquid
chemicals at a constant rate of approximately 50 gal/min. To apply
different amounts of liquid  chemical in the different soil zones,
the time to penetrate the soil at different depths is varied. The drill
operator controls the penetration of the augers into the soils, and
a depth gage is monitored by the operations director.  The rate and
quantity of liquids injected can be monitored with rate and total
flow meters  for each auger.
  The liquid chemicals are injected through the kellys to the auger
bits and are ejected from the liquid/steam jets on the augers. The
pressure of the ejected stream and the rotation of the  augers mixes
the liquid chemicals into the soils. Advancement of the augers in
the loosened soils after the initial intrusion is easily accomplished.
  The time required for the liquid chemical injection operation
is  dependent upon the prescribed amount  of chemicals to be
injected. The more highly contaminated zones require more chemi-
cals and time because of the fixed rate  of liquid  chemical feed.
Experience at the site indicates that typical liquid chemical injec-
tion requires 8 min and 13 min for 15 ft and 22.5 ft of penetra-
tion, respectively.

Stripping Operations
  The hot air and steam are injected into the soil block through
the auger bits as the augers  rotate and are raised or lowered through
the contaminated soil zones. The mixing action of the augers assists
in exposing the soils and contaminants to the  steam and hot air
for stripping. Normal rotation speeds are 10 and 20 rpm for the
bits, and the bits are raised slowly from  the deeper zones  toward
the surface at approximately 2 ft/min. Traversing back to the
bottom of the soil block is normally rapid, at  a rate of approxi-
mately 10 ft/min.
  The soil  block is steam  and hot air stripped as the operations
director monitors the concentration of hydrocarbons in the  off-
gases,  the position and rate of advancement of the augers  and the
elapsed time. Instructions are conveyed to the tractor operator who
controls the drilling operations.
  Varying amounts of time and numbers of traverses of  the soil
block have been attempted in stripping  operations.  In  addition,
some experimentation with the concentrations in the off-gases and
the amount of  stripping  has  been attempted. The amount of
stripping has ranged from about 10 min upwards to 2.5 hr. Normal
stripping times prescribed for this site are 30 min for each soil block,

                           Table 1
                    Soil Contamination Data
Soil

Row
B
6
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
C
C
C
C
C
C
C
C
0
D
D
D
0
D
D
H
H
H
H
H
H
I
I
I
I
K
K
T
T
M
Notes

Block

No.
3
3
4
4
5
5
5
7
7
B
8
9
9
10
10
11
11
7
7
8
e
9
9
10
10
4
4
5
5
10
10
10
e
6
6
7
7
7
6
6
8
8
7
8
8
8
8
: "d"
•(...:
Depth Totil

ULl
5
10
5
10
1
5
10
5
10
5
10
S
10
5
10
5
10
5
10
5
10
5
10
5
10
5
10
5
10
1
S
10
1
5
10
1
S
10
5
10
S
10
10
S
21
21
21
- dlesel fuel
1" • additional test
Hydrocarbon Content
Before
Treatment
2700
250
880
480

1300
270
160
240
510
790
1000
310
300
430
2100
23
54
110
320
970
470
310
330
1200
1200
320
800(900d)
440

74
4040
eeo
460
325
260
360
1100
170
800
390
770
340(lll)(190d)
170d
15000
28000
6600
results
(ppm)
After
Treatment
130 (43d)
130
77
71
10
66
270
180
150
160
100
110
69
150
110
92(<25d)
110
86(81)
51
68
43
120
68
120
130ซ25d)
120
64
250(54d)
76
87
110
120
150
120
85
ISO
130
100
240
130
140
100
31d
33d
160(2200) (110)
1300(4700)(130)(61)
480(1700)

                                                                                 CONTAMINATED SOIL TREATMENT     401

-------
 and it is performed in 3 to 6 traverses of the augers through the
 soil block.

 Soil Stabilization
   After the stripping operations are completed, a stabilization oper-
 ation  can be performed on the soil block which completes the
 operations. The loosened and moist soil block can be stabilized
 by the addition of dry reagents into the soil block. Dry cement
 was metered  into the hot air return conduits which convey the
 cement through  the kellys and discharge it from the port at the
 auger bits. The rotating augers mix the cement into the soil mass
 as the augers traverse through the soil block from top to bottom
 and return.
   The operations director monitors the feed rate of cement and
 continues  the injection, mixing and traversing operations until the
 prescribed amount of cement has been placed in the soil mass. A
 typical amount of cement to be added to a  16.5-yd3 soil block was
 629 Ib distributed through the 15 feet depth of penetration. This
 was accomplished in about  12 min. Some additional time prob-
 ably was required to ready the equipment for this operation which
 can be controlled from the  control room.

 Coupled Processes
   Coupled processes  which are performed while the in situ treat-
 ment operations are performed, consist of:
 • Off-gas condensation
 • Moisture removal (demisters)
 • Carbon adsorption
 • Return air heating
 • Return air pressurization
 • Superheated steam generation

   These coupled processes occur simultaneously with the other
 operations. These processes condition the off-gases for return to
 the soil block  for more stripping of hydrocarbons. No additional
 time normally is required for these operations.
Treatment Time
  The time required to perform the treatment operations on a soil
block varied from 10 min for very lightly contaminated soil blocks
to over 2.5 hr for some blocks which were extensively stripped.
For the prescribed 30 min stripping time for this site,  the time
required for the in situ treatment operations and movement between
blocks was about 45 min per soil block.

SOIL TREATMENT RESULTS
  Data were obtained at the site to demonstrate the in situ  treat-
ment process. These data included chemical tests on soil samples
obtained before and after the treatment process and logs and charts
of the treatment operations on the soil blocks. These data will be
described  in this section.

Soil Sample Tests
  Soil samples were obtained at various depths in the location of
the soil blocks  before  the treatment  process was  employed.
Undisturbed, "Shelby tube," soil samples were obtained by con-
ventional soil sampling methods. Samples were obtained at depths
of 1, 5, 10 and 20 ft as prescribed.
  The samples were tested for total hydrocarbons, benzene, diesel
fuel and gasoline. The samples were tested for total hydrocarbons
by U.S. EPA Method 413.2.  In addition, some samples were
extracted and tested by gas chromatography according to U.S. EPA
Method 8015. After the soil block was treated by the in situ treat-
ment system, the soils were again sampled.
  The results of the analytical tests indicate that the total petro-
leum hydrocarbons were reduced substantially, with reduction of
up to 95%. Initial concentrations in the soils were as high as 40,000
ppm. the remedial operations successfully reduced the overall con-
centrations of hydrocarbon fuels  to levels below the cleanup goal
of 100 ppm. A summary of some of the soil test results is presented
in Table 1. Reduction of the concentration of hydrocarbon fuels
ranged up to 95% with one treatment  on  the in situ treatment
system.
402    CONTAMINATED SOIL TREATMENT

-------
                                        Mixed Waste  and  Alara

                                                    Nancy P.  Kirner
                                                Envirosphere Company
                                                 Bellevue, Washington
ABSTRACT
  Mixed radioactive and hazardous wastes only recently have been
recognized as a classification of low-level waste deserving special
attention. This paper defines low-level mixed wastes, discusses some
of the problems encountered in managing low-level wastes and
provides an example of how two state agencies balanced the need
to comply  with  RCRA with the need  to  maintain radiation
exposures to levels which are as low as reasonably achievable. A
potential solution to the current mixed waste disposal problem is
presented.

WHAT IS MIXED WASTE?
  Low-Level Radioactive Mixed Waste (mixed waste) is low-level
radioactive wastes  which also contains  materials  regulated as
hazardous waste under RCRA1. Low-Level radioactive waste is
defined in a convoluted manner in the Low-Level  Radioactive
Waste Policy Amendments Act (LLRWPAA) to be Atomic Energy
Act materials which are not spent fuel, not  wastes from repro-
cessing, not uranium mill  tailings, not  transuranic  wastes, but
materials which are considered to be low-level waste according to
the NRC. RCRA regulated materials are those which are specifi-
cally listed in Subpart D of Title 40 CFR Part 261 (40 CFR 261)
or which exhibit the hazardous characteristics, defined in Subpart
C of 40 CFR 261, of ignitability, corrosivity,  reactivity or extrac-
tion  procedure toxicity.
  The definition of mixed waste can vary depending on the regula-
tory  authorities involved. For instance, the State of Washington
is an Agreement State, with delegated authority from the NRC to
regulate low-level radioactive waste and delegated authority from
the U.S.  EPA for the  regulation of hazardous waste. The State
of Washington also has some delegation of regulatory authority
from the U.S. EPA, specifically for mixed waste. In addition to
these federally delegated programs, the state also regulates materials
under strictly state authority. Naturally occurring and accelerator
produced materials are examples of state regulated radioactive
waste.  The State  declares as hazardous certain materials not
regulated under RCRA. In general, the State of Washington does
not consider how the material was generated as much as it con-
siders how the disposal of the material may affect the environment.
This  appears to be a potentially important distinction when the
disposal of lead as shielding is considered.
  The NRC and the U.S. EPA consider the regulation of mixed
waste to be "dual regulation," with the disposal requirements for
the radioactive component added to the disposal requirements for
hazardous components of the waste.  Dual regulation of mixed
waste disposal,  however,  could involve multiple separate state
regulatory agencies (possibly one radioactive and one hazardous
regulatory agency at both the  point of generation and the point
of disposal, with additional agencies for any intermediate waste
treatment facilities) as  well as the two  federal oversight agencies.
Thus, the classification and disposal of mixed waste is not a stream-
lined process if one merely considers the regulatory agencies and
authorities involved in determining what constitutes mixed waste.

WHO GENERATES MIXED WASTE?
  Based  on a survey of shipping  papers,  Bowerman, et al.2
reported: most generators who identified chemical constituents on
the radioactive waste manifest were non-fuel cycle generators;
mixed wastes comprised no more than 5% of the total volume of
waste arriving at Hanford; and most of the waste resulted  from
liquid scintillation counting techniques. The report also stated that
a radioactive shipment manifest does not contain sufficient infor-
mation to conclusively determine whether a material, even if it were
listed on the manifest, was subject to the disposal requirements
of RCRA. For that reason, the shipping manifest survey did not
conclusively determine the extent of mixed waste disposal.
  Another survey was performed using a questionnaire technique.
Representative generators of low-level radioactive waste were asked
to identify wastes which they considered to be subject to RCRA.2
Organic liquids, primarily those used in scintillation counting, lead
and chromates used as corrosion inhibitors, were identified in the
survey as the primary sources of mixed waste. Quantities of mixed
waste were similar to the previous study's  results. One problem
existed with both of these studies: they relied heavily on the gener-
ator's knowledge of RCRA which may not have been accurate.
  My personal experience  with waste generated  in a university
setting confirms the Bowerman reports.  Liquid scintillation
counting represented the largest volume of mixed waste. Biomedical
researchers' and suppliers' contributions may be underestimated,
however. This sector utilizes a myriad of organic solvents, drugs
and drug precursors, acids and bases and chromate-based cleaning
compounds, which, if untreated, could require RCRA) regulated
disposal. Some industrial sources were overlooked in the reports.
At least one manufacturer of mercury vapor lamps attempted to
dispose of elemental mercury and strips of thorium coated metal.
Some of the mercury leaked out of the disposal package, alerting
agencies to the ever present potential for mixed waste disposal.

PRESENT OPTIONS FOR DISPOSAL OF MIXED WASTE
  Presently, no  commercial low-level waste disposal  site will
knowingly accept mixed waste for disposal. Organic liquids could
and, under RCRA,  should be treated by incineration, but no
incinerator is commercially available for NRC regulated liquids,
those which fall  outside of the NRC biomedical rule, 10  CFR
20.306. Thus, regulated mixed waste must be stored until disposal
options are available. Storage, itself, is a dually regulated activity
under the Atomic Energy Act and RCRA.

LEAD
  Lead and heavy metals represent a potentially significant source
of mixed waste, not only because of their volume, but also because
of their ideal radiation shielding characteristics. Lead was recently
                                                                                        RAD AND MIXED WASTES    403

-------
the subject of an interpretive letter from Marcia Williams, Direc-
tor of Solid Waste at the U.S. EPA3. Lead mixed waste can be
classified into three categories: surface contaminated; intrinsically
contaminated or activated; and shielding for radioactive waste
which will continue to shield the waste after disposal. Some tech-
niques are available to decontaminate surface contaminated lead,
either by physically removing the radioactive material or by smelting
and separating the lead from its radioactive impurities. Lead from
these processes may be recycled and the radioactive contaminants
saved for eventual treatment and/or disposal as mixed waste. Lead
which is intrinsically contaminated or activated presents a more
difficult management problem which may not lend itself to smelting
or other traditional techniques. Lead as shielding, according to the
Williams letter, is not subject to RCRA as RCRA does not extend
to packages and the lead is not a waste as it continues to serve its
intended purposes even during disposal.
  The State of Washington has informally taken exception to these
distinctions as the same potential for environmental impairment
exists whether or not the lead is serving a useful  purpose in the
disposal site. The Williams letter also allowed for the possibility
of encapsulating mixed waste lead in a "stable" material, provided
the waste could pass the EP or TCLP toxicity test. Although the
letter was ambiguous as to  encapsulation materials and testing
methods, it provided some needed practical thinking on ways to
meet RCRA requirements without violating  the basic radiation
protection  philosophy of maintaining exposures to radiation to
levels as low as reasonably achievable (ALARA).

WHY IS ALARA IMPORTANT?
  The concept of maintaining radiation  exposures ALARA was
adopted by the old Federal Radiation Council, a precursor agency
to the U.S. EPA, in an effort to balance trade-offs and risks asso-
ciated with a postulated non-threshold dose/effect  relationship,
the need to take protective measures even if regulatory limits may
be  met and  the costs of implementing those extra protective
measures. ALARA becomes important in the extent of Section 1006
of RCRA.  This section could be interpreted as allowing Atomic
Energy Act (AEA) requirements, such as the ALARA concept, to
take precedence over RCRA requirements if RCRA requirements
are inconsistent with Atomic Energy Act requirements.  Inconsis-
tency has been postulated  to mean: an increase in the radiation
hazard, technical infeasibility or a violation of national security
interests.
  Both NRC and the U.S. EPA have been engaged in a detailed
review of RCRA and AEA disposal requirements and have deter-
mined that, on a conceptual basis, there do not appear to be any
inconsistencies between RCRA and AEA requirements. However,
it is possible that questions of inconsistency could  be raised when
dealing  with a specific waste.  For  instance, is the  ALARA
philosophy implemented when pre-packaged  waste is sampled at
a disposal site to verify its contents (per  RCRA requirements) if
an inspection could be performed more effectively and with less
hazard at a generator's facility?
  A recent case of balancing the ALARA concept with the require-
ments of RCRA occurred  at a waste disposal site where under-
ground tanks were used to treat liquids by solar evaporation. The
solar evaporator never worked leaving as its legacy  five  tanks of
unknown size, containing unknown liquids and sludges, with
incomplete or inadequate records. The tanks had been  sampled
and their radioactive constituents characterized several years prior
to a concern over the tanks containing mixed waste. Some tanks
represented occupational dose rates in excess of 10 roentgens/hr.
With new tank restrictions and new mixed waste issues  looming,
how much, if any, further sampling could be  considered to be
ALARA?
  After much discussion and compromise between the involved
state agencies, it was decided to impose a budget of 1.25 man-rem
(whole body) and a 10 extremity-rem (maximum exposed extremity)
for the collection and initial processing of representative samples
from all five tanks. If the budget were exceeded before all five tanks
could be  sampled to the extent desired  for characterization,
sampling would cease and the contents of uncharacterized tanks
would be considered to be subject to RCRA. It was in everyone's
best interest to collect as many samples with as little dose as pos-
sible. Strict radiation control measures were employed with the end
result being five tanks adequately characterized with only a frac-
tion of the allowed dose  actually received.
  This case illustrates how the two sets of requirements can be met
while taking into consideration ALARA concepts. One source of
difficulty is that there are few numerical values as to what consti-
tutes ALARA. Modifications are required  if a man-rem is saved
by spending $1,000 or less in 10 CFR 50. However, this guidance
is not specific to situations other than  nuclear power plants, nor
does it reflect current economic conditions. It therefore appears
unlikely that  the U.S. EPA or NRC will support any determina-
tion of inconsistency with  RCRA based on a potential conflict with
the ALARA concept, unless it can be demonstrated that signifi-
cant  and certain harm will result.

CONCLUSION
  Because of the current  unavailability of commercial waste dis-
posal and treatment options, elimination or minimization of mixed
waste generation appears to be the preferred alternative for both
the short- and long-term.  This may be accomplished by primarily
technical advances such as product substitution; source control by
material segregation, training and process modification; recycling;
and in-process treatment.
  Institutional solutions  must also be sought. Recognizing that
elimination of all mixed waste cannot be accomplished,  one
immediate possibility is the de-listing by the U.S. EPA of specific
waste streams from a specific plant or industry.
  Ultimately  needed are a streamlined and coordinated regulatory
system and a  set of regulatory requirements specifically designed
to meet the total hazard which each mixed waste represents. This
system will require a more in-depth understanding of the types,
quantities and chemical characteristics of the mixed wastes being
generated presently. Such a coordinated regulatory system would
not only take into account the total hazard of the waste, but also
would allow wastes having a predominate hazard to be disposed
following requirements  for  only the  predominate hazard4. De
minimis disposal concepts should be applied to both radioactive
and hazardous wastes. An analogous situation to the present system
of dual regulation of mixed waste may be the making of a ham
and cheese sandwich by combining a ham sandwich with a cheese
sandwich. Instead of combining the ham and cheese together into
one unified, coordinated sandwich, the current regulatory structure
would require four slices of bread, twice as much butter or mayon-
naise and additional expense.

REFERENCES
1.  Joint EPA/NRC Guidance on the Definition and Identification of Com-
   mercial Mixed Low-Level Radioactive and Hazardous Waste, Jan. 8,
   1987.
2.  Bowerman, B.S., Kempt, C.R.,  Mackenzie, D.R.,  Siskind,  B. and
   Piciulo, P.L. 1985 "An Analysis of Low-Level Wastes:  Review of
   Hazardous Waste Regulations and Identification of Radioactive Mixed
Wastes," NUREG/CR-4406, U.S. Nuclear Regulatory Commission.
3.  Letter from Marcia E. Williams, Director, Office of Solid Waste,  U.S.
   EPA, Washington, D.C. 2-46- to Mr. Terry Husseman, Washington
   Department of Ecology,  Olympia, WA, June 26,  1987.
4.  Augustine, R.J., "Mixed Waste: A Generator's Perspective", paper deli-
   vered at the 19th Annual National Conference of Radiation Control,
   Boise, Idaho, May 1987.
404    RAD AND MIXED WASTES

-------
               Overview  of the  West Valley  Demonstration Project

                                                    M.  D.  Weingart
                                                 Ebasco  Services, Inc.
                                                    New  York,  NY
                                           R. R. Borisch and D.  R. Leap
                                      West Valley  Nuclear Services Co.,  Inc.
                                                   West  Valley,  NY
ABSTRACT
  The West Valley Demonstration Project (WVDP) was authorized
by Congress in 1980 through the West Valley Demonstration Act.
This legislation  requires the  Department of Energy (DOE) to
solidify the more than 2120 m3 of liquid high-level nuclear waste
currently stored underground at this former commercial nuclear
fuel reprocessing facility located near Buffalo, New  York. This
solidification product eventually will be sent to the federal high-
level waste repository.
  The technology selected for processing this waste is vitrification
(encapsulation in glass). A 47.2-Mg, ceramic lined, glass melter
is the principal feature of the vitrification system. The melter along
with temporary  support systems is installed and presently  is
undergoing nonradioactive testing utilizing chemically similar sub-
stances to simulate the waste and to characterize the final glass
product. This testing will continue for several years during which
time the facilities will be upgraded for radioactive operation.
  This paper describes the initial condition of the site and facili-
ties,  some of the activities (such as decontamination) which are
being conducted to allow use of existing facilities and some of the
problems encountered during construction (such as contaminated
soil and existing underground interferences). Also  discussed are
the various systems and facilities associated with the vitrification
process including the  Vitrification Facility, High-Level Waste
Pretreatment  System and Facility, Cold Chemical System and
Facility, Low-Level Waste Treatment System and Facilities, Sludge
Mobilization System and the plans for waste package interim
storage and shipout to the federal repository. The present Project
engineering and construction status will be addressed along with
the Project schedule leading to radioactive operation.

INTRODUCTION
  The  Western  New York Nuclear Services Center near West
Valley, owned by the New York State Energy Research and
Development Authority (NYSERDA), was the site of the only com-
mercial nuclear  fuel reprocessing facility ever to operate in the
United States. Nuclear Fuel Services Company (NFS), a subsidiary
of Getty Oil, was the facility operator. The plant reprocessed spent
nuclear fuel assemblies from various nuclear power plan s from
1966 to 1972.  Reprocessing operations generated approximately
2120 m of highly radioactive liquid waste. When the plant was built
in the early 1960s, the approved method of high-level  radioactive
liquid waste storage was underground storage in steel tanks.
  NFS closed the plant in 1972 in order to expand it. However,
increased Federal and State regulations applicable not only to the
planned expansion but  to the entire facility made the  required
capital investment much more costly than had been anticipated,
and NFS decided not to proceed with the project. The site was
turned back to NYSERDA, leaving the liquid waste in temporary
storage tanks at the site.

West Valley Demonstration Project Act
  In 1980,  the United States Congress passed the West  Valley
Demonstration Project Act (PL96-368) authorizing the DOE to
carry out a nuclear waste management project at West Valley. The
Project was established to demonstrate that liquid waste from
reprocessing of spent commercial nuclear fuel can be managed
safely in the United States.

Project Approach
  The DOE assumed control of the West Valley site in February
1982. West Valley Nuclear Services Company (WVNS), a division
of Westinghouse Electric Corporation, was chosen as the site prime
contractor.  During early Project planning, it was determined that
since the technology base for waste solidification was well known,
the best "demonstration" at West Valley would be the expeditious
adaptation  of existing technology. An  aggressive and creative
"Actiontrak" approach was adopted  which scheduled detailed
design and implemented construction and decontamination activi-
ties to be conducted essentially in parallel. Ebasco Services, Inc.
was chosen as the Architect Engineer to perform major Project
design/engineering.
  The Project would make the maximum feasible use of existing
on-site facilities. This would minimize new construction, demon-
strate the feasibility to reuse former decontaminated nuclear process
facilities, and reduce the extent of the environmental stabilization
effort required at  the end of the Project.

Initial Site Conditions
  Following site takeover in February 1982,  WVNS began an
extensive program of site environmental characterization and plant
refurbishment. In anticipation of the future solidification Project,
remodeling  in the plant and administration building was begun to
provide additional office space.
  All older radiation detection equipment was replaced with the
latest available equipment. Chemical analysis laboratories were
renovated and new analysis equipment was added. The remote
analytical lab aisle windows were cleaned of chemical deposits and
remote handling equipment and cranes were refurbished. Decon-
tamination  of the former reprocessing  plant was begun. Over
3577 m2 of space were cleaned the first year.
  Gamma surveys were performed for over 371,600 m2 of fenced
protected area, and over 120 soil and vegetable samples were taken
                                                                                       RAD AND MIXED WASTES     405

-------
to establish the background radiation baseline. Existing USGS on-
site monitoring wells were sampled for radiological and chemical
analysis. Additional USGS  test wells  were drilled  around the
perimeter of the Nuclear Regulatory Commission's licensed waste
disposal area to establish the geologic information in that area.

Hlgh-Level Waste Tank
  The majority of the high-level liquid waste is stored in an under-
ground  steel tank (Figure 1). During sampling of the high-level
liquid waste, Project engineers discovered that 90% of this waste
is a supernatant  liquid and  contains Cesium-137 and traces of
Strontium-90 and barium. The remaining  10% of the high-level
liquid waste is in the form of sludge at the bottom of the tank.
Strontium-90 is the major source of radioactivity in  the sludge.
  The high-level waste tank is made from carbon steel. It  rests on
a steel saucer on a perlite block pad and is enclosed in a concrete
vault. Ultrasonic and photographic inspections have confirmed the
tank's 12.7-mm thick plate walls are sound. Special equipment has
been designed to aid in removing the supernatant and sludge from
the tank.
  The supernatant liquid will  be pretreated to remove cesium which
is the primary radioactive fission product. The resulting low-level
liquid will be processed in the low-level waste treatment facility
and then will be capsulated in  cement. Other low-level waste streams
resulting from plant operations will be concentrated and processed
into cement, also. The radioactive cesium stripped  from the super-
natant will be recombined with the sludge  and ultimately will be
made into a permanent insoluble glass.
   WAVCl WATER
   INJECTION UNf
        fTEEL tMICflt

                 FERUTC BLOCK*
                           Figure I
                 High-Level Liquid Waste Tank
Hlgh-Level Waste Solidification
  The major equipment necessary for making glass has been in-
stalled in the vitrification facility and is operational (Figure 2). The
basic vitrification system consists of a 47.2-Mg joule heated ceramic
melter, a large mixing tank named the Concentrator Feed Make-
Up Tank (CFMT) and an off-gas purification system. Test glass
making runs are 35% complete in preparation for radioactive oper-
ations. These runs are gathering important statistical information
from a sophisticated computer system that  also will be used to
remotely control process operations.
  The melter's refractory brick is designed to accept waste streams
of different chemical compositions. Liquids are evaporated under
intense heat allowing glass formers and waste to be dissolved, atom-
ically bonding them together. The molten liquid is made homo-
geneous through the natural process of convection. Gases from
the melter are cooled and filtered in the off-gas system. Clean gases
are  discharged to the environment  through a  monitored high
efficiency off-gas filter system.
                          Figure 2
                     Vitrification Facility
Vitrification Facility
  The building housing the vitrification test system was cc
iplftfd
in 1984. The building was designed to accommodate the testing
of the major vitrification components in parallel with construc-
tion of the final shielded cell. Construction of the facility begin
in 1983. Excavation for the below grade structure unearthed some
contaminated soil. This soil had to be removed and boxed before
work  could continue. The construction program was accelerated
during 1984 to ensure that the below grade shielded cell pit and
Butler-type building structure were completed to allow the early
delivery and  installation of the melter and canister turntable.
  Innovative thinking by the WVNS Construction Department per-
mitted extensive concrete pouring during the winter months. This
teamwork resulted in the first glass log being poured in December
1984,  over 6 months ahead of schedule. Construction of 1.2-m thick
shield walls to enclose the melter and support equipment is approxi-
mately 50% complete. Modular concrete wall panels have been
developed to allow for equipment testing prior to completion of
the construction of the cell walls. All of these efforts are being
conducted to support start of high-level waste vitrification in arty
1991.
Hlgh-Level Waste Prefreatmeal
  The supernatant will be pumped from the high-level waste tank
to a liquid waste treatment system installed within an existing spire
underground waste tank. The Supernatant Treatment System (STS)
main  components are a series of ion exchange columns that will
use zeolite ion exchange media (a clay-like granular material) to
extract radioactive cesium from the liquid (Figure 3). This spare
tank  provided  the necessary  shielding for  four Jon-exchange
columns which are approximately  6.1  m  high and 0.91 m in
diameter. The  system is designed  to move  liquid sequentially
through three columns to complete the extraction process. The
fourth column will be off-line  having the loaded zeolite removed
and replaced with fresh zeolite. When the first column in the series
 iHoni
 Will*
                            Figure 3
              West Valley HLW Processing Flow Sheet
406     RAD AND MIXED WASTES

-------
no longer effectively removes cesium, a series of valves are acti-
vated to rotate the flow sequence so that the first column conies
off-line, the second and third columns move up one position and
the fourth column comes on-line in the number three position. The
system is set to make this sequence change without interrupting
process flow. The used zeolite will be collected on the bottom of
the spare tank until all the supernatant has been processed. Pre
and postfilteration of the waste stream are provided. Construc-
tion and installation of the STS major vessels and equipment has
been completed.
  A valve aisle and control building were erected adjacent to and
above the spare waste tank. Specially designed pilings were put
in place near the tank to support the control building and valve
aisle floor. A remote installation system designed in cooperation
with the Rockwell Hanford Company was used to remotely install
risers in the high level-waste tank. This operation was necessary
since there were not enough openings in the tank or vault top to
properly configure the sluicing pumps which are required  to
mobilize the sludge in the bottom of the high-level waste tank. Five
similar risers were required in the spare waste tank to remove the
spent zeolite. These risers were installed first in the spare tank as
practice prior to moving over and installing the risers on the high-
level  tank. A  high pressure water/grit cutting system called  an
"Admac" was used to cut the 0.61 m thick existing concrete  vault
top to inset the Supernatant Treatment System, pumps and tanks
into the spare tank. The spare tank was sufficiently radioactively
contaminated inside that all tank modifications and STS instal-
lation were performed without ever entering  the tank.

Sludge Mobilization
   Special sluicing pumps have been designed to mobilize the sludge
settled the internal grid  structure on the bottom of the high-level
waste tank. The grid structure presents unique challenges  since
sludge solids have settled in the structure (Figure 1) of 0.23 m
I-Beams and 0.30 m girder plates. A one-sixth scale model of the
tank was used to design the mobilization pumps and demonstrate
that the sludge can be removed. Zeolite from the STS will be mixed
with the sludge and transferred to the melter for vitrification.

Cold Chemical System
   During radioactive operations, the Cold Chemical System (CSS)
will ensure  the HLW glass log quality and uniformity by  strict
control of the non-radioactive chemical additives. During cold
testing, this system will also be used to provide a synthetic waste
similar to that found in the HLW tank.
   The HLW glass logs will consist  of radioactive waste,  glass
formers and "surrogate waste simulants." The amount of waste
simulants will be adjusted to compensate for variations in the HLW
streams coming from the tank. The waste simulants will maintain
a uniform waste to glass  ratio.
   This system is capable of batch preparation of up to 11  glass
formers and 37 waste simulants. Each batch can be custom mixed,
pulverized and transferred to the hot cell during radioactive  oper-
ations. The system will also provide for preparation of special
decontamination solutions for cleaning process equipment in need
of service during vitrification operations as well as during vitri-
fication facility decommissioning.

Low-Level Waste Treatment Systems and Facilities
  Once decontaminated, the supernatant liquid will be transferred
through underground piping to the Liquid Waste Treatment system
(LWTS) located in the former reprocessing facility. The LWTS
has been installed in  former chemical process cells which were
decontaminated and refurbished for this purpose.
  The LWTS (Figure 4) will filter, demineralize and concentrate
(via  evaporation) this waste. Other facility  waste  also will  be
processed in the system. Spent resins, filter backwash and other
Class B and C liquid plant waste also can be solidified in cement.

Cement Solidification System
  Installed in the  former  reprocessing facility,  CSS  became
operational in  1985. The system  uses  high shear mixers  to
thoroughly blend waste streams with Portland cement (Figure 4).
Waste form qualification tests have certified that recipes developed
at West Valley meet the requirements of 10 CFR 61.
  Approximately 12,113-1 of uranyl nitrate hexahydrate have been
successfully solidified into cement using 208-1 round drums as con-
tainers. A series of conveyors move empty drums through shield
walls to a cement injection station. Drums are filled with cement
and  capped.
  The CSS was taken out of service in 1986  for refurbishment.
It has been redesigned to accept 269-1 square drums in addition
to 208-1 drums. Once supernatant  processing begins, the STS,
LWTS, and CSS will work as an integrated system. When super-
natant processing is completed, approximately 13,000 269-1 drums
will  have been processed.

High-Level Waste Storage
  Approximately 300 glass canisters will be produced during the
solidification program. The molten glass will be poured into 0.61-m
diameter by 3.0 m long stainless steel canisters. A large mechani-
cal turntable has been designed to receive four canisters. The turn-
table can be rotated to allow  continuous glass pour once the
program begins.  It takes approximately 36 hr to fill a canister.
  Once the high-level waste has been solidified, the canistered waste
glass will be remotely transferred to  interim storage in the former
reprocessing cell known as the Chemical Process Cell (CPC). This
facility will be modified to hold the 300 canisters prior to their trans-
fer to a federal repository for permanent disposal.

Dram Cell
  Drums established by assay to have less than 100 nCi/g of tran-
suranic elements  will be placed in a monitored corrugated steel
building (Figure 5). Remotely operated cranes will be used to stack
drums on a specially designed clay and gravel pad. The pad pro-
vides an engineered drainage system preventing water from reaching
the drums. If waste disposal at this  location on the site is viable,
                           Figure 4
      Radwaste Treatment System (RTS) Processing Flow Sheet
                           Figure 5
                        RTS Drum Cell
                                                                                            RAD AND MIXED WASTES    407

-------
 this building will be converted to an earthen structure called a
 tumulus (Figure 6). Low-level wastes would be permanently isolated
 from the environment using a series of cement, clay and geotextile
 barriers. The pad, cement footers and building structure have been
 completed. The building will be ready to  receive drums in Sep-
 tember 1987.
                         True Scale
Compacted Clay Liner
Shield Wall	
                                       31 Slope
Qeolexllle Reinforcing

Compacted Clay
Drain & Intruder
Barrier
Ground Surface
Waile Boxel
 Watle Drumt
                                               ซ-^—Oravel Drain
                                                   "-Oeolexllle
                  Vertically Exaggerated
                                      Compacted Clay
                            Figure 6
              Schematic of Above Grade Disposal Unit
 Decontamination and Decommissioning
   The former nuclear reprocessing plant is comprised of a series
 of sealed concrete cells. The majority of these cells are tall and
 narrow, with tight, hard to reach spaces. Entrance can be made
 only through a top hatchway or bottom door. Over 640 tons of
 nuclear fuel was reprocessed in these cells using remotely operated
 equipment  and  mechanical  manipulators.  Shielded  viewing
 windows were utilized to observe the chemical process and perform
 equipment  maintenance.
   Significant  decontamination  and  decommissioning (D&D)
 experience has been gained at West Valley.  Prior to cleanup, each
 cell was characterized using remote  video cameras and contami-
 nation surveys. Exact locations of contamination and radiation
 dose levels  were known prior to manned entry.
   To date,  over 70% of the plant is now ready for reuse (Figures
 7, 8). Over  63 Mg of tanks and vessels and 5.79 km of pipe have
 been removed from the eight cells necessary for the LWTS. Also
 nearing completion is cleanup of the former Chemical Process Cell.
 27.4 m 7.9 m wide and 12.2 m high, this is the  largest cell at West
 Valley. It will house the glass canisters until a federal repository
 is ready. Over 159 Mg of tanks and vessels have  been removed from
 this cell.
  Another  major part of the D&D  program  was the  successful
 shipout of  utility  owned  spent  fuel. Six hundred twenty-five
 assemblies owned by General Public Utilities, Rochester Gas &
          Total Calculated Squara Foolaga Ol Araa-350.000
                          Squara Feat
         FYB2 WVNS Tlkeovir
                                      Slilui at ol Jinuary 1987
                           Figure 7
              Total Facility Decontamination Status

408     RAD AND MIXED WASTES
                 Electric, Wisconsin Electric Power Company and Commonwealth
                 Edison Company were safely returned. Remaining are 125 assem-
                 blies now owned by the Department of Energy. These assemblies
                 will be used to demonstrate dry cask storage at another DOE
                 facility. The former fuel pool will be used for underwater cutting
                 of tanks, vessels and piping removed  from the facility.

                 Project Schedule
                   The present project schedule targets the spring of 1991  for the
                 first high-level glass canister. Present engineering/design,  construc-
                 tion and startup efforts are proceeding in a manner which supports
                 this date.
                                                                                                              PCR
                                                                ALPHA LAB

                                                                ANA Call
                                                                  URAN LAB
                                                                                                               k frearete
                                                                                                               hi Ptefrtit
                                             Figure 8
                                  WVDP Decontamination Activity

-------
              Application  of  the Remedial  Action  Priority  System
                              To Hazardous Waste Sites  on  the
                                       National  Priorities  List

                                                    Gene Whelan
                                               Robert D.  Brockhaus
                                                 Dennis L. Strenge
                                              James  G. Droppo, Jr.
                                                 Marcia B. Walter
                                                   John  W. Buck
                                          Pacific Northwest Laboratory
                                              Richland, Washington
ABSTRACT
  To address the intent of SARA for the development of a "revised
Hazard Ranking System (HRS)" the U.S. EPA is reviewing and
testing several ranking methodologies to determine the most appro-
priate one to represent the revised HRS, among them the Remedial
Action Priority System (RAPS). The application of the RAPS
methodology to a number of hazardous waste sites on the National
Priorities List (each independently chosen by the U.S. EPA) each
site's rankings and HRS scores where available, are discussed. More
importantly, this paper identifies (as determined by RAPS) the con-
stituents of major concern, principal transport pathways and
exposure routes and, specifically, why each site received its score.

INTRODUCTION
  With the passage of RCRA CERCLA and SARA' Congress has
mandated a much closer scrutiny of management of hazardous
wastes generated now and in the future and restoration of disposal
sites contaminated through improper management. Legislative
language, regulatory intent, and prudent judgment call for risk
assessment techniques  to aid in the decision-making process; this
applies to both publicly and privately operated facilities. This con-
gressional mandate is especially exemplified in SARA. One example
is the rule to "promulgate amendments to the hazard ranking
system" such that it "accurately assesses the relative degree of risk
to human health and the environment posed by sites and facilities
subject to review."
  Several screening methodologies separate  potential environ-
mental problems into two categories:  those that may pose a sig-
nificant hazard to the surrounding environment and those that pose
little or no hazard. However, these methodologies cannot priori-
tize environmental problems and generally do  not base their
rankings on relative risk. Before the passage of SARA' the U.S.
Department of Energy's Office of Environment, Safety and Health
supported the development of  the RAPS system to assess the
relative degree of risk to human health and the environment.
  RAPS is an objective, physics-based' fully integrated multimedia
environmental assessment system that ranks hazardous and radio-
active mixed-waste sites based on relative human health risk and
limited site information. Site information can  be obtained from:

• Site inspection
• Site documents
• Local, state, or federal reports and documents
• Measured/monitored data (i.e.. information that can only be
  supplied by sampling at the site)
• A data base developed  specifically  for RAPS
  RAPS uses site-specific data when available, local or regional
data where appropriate and default values when necessary.
  RAPS  is a  computer-based system that uses  empirically,
analytically and semi-analytically based mathematical algorithms
to project the release of contaminants into the environment, con-
taminant transport through and between multiple environmental
media (i.e.,  air,  groundwater, surface  water  and overland),
exposure  to surrounding populations  and health effects (i.e.,
relative risk) associated with exposure. Based on the exposure and
health effects components, RAPS calculates  a Hazard Potential
Index (HPI) representing the parameter used to rank sites.  The
HPI, although open ended, generally ranges from 0 to  100 with
its scoring system unrelated to that of the HRS. RAPS is intended
to bridge the technology gap between initial  site evaluation and
the time-consuming process of actual field site characterization,
assessment and remediation. Comprehensive reviews are provided
by Whelan, etal.1-2-3.
  This paper illustrates application of the RAPS methodology to
several hazardous waste sites. RAPS was applied to 20 CERCLA
sites; for  brevity, this paper  discusses  nine of the 20 sites. The
rankings scored by each site are reviewed and  HRS scores are also
provided, where applicable. Finally, a brief review of each site
presents the major constituents of concern; dominant transport
pathways and exposure  routes; and reasons behind each site's
ranking.
  Each site review is based on site conditions identified by U.S.
EPA before  the RAPS  analysis and do not necessarily reflect
current conditions. For example some form of remedial action was
performed at several of the sites, but these activities were not in-
cluded in the assessment, and the reviews only reflect pre-remedial
conditions, (although RAPS can handle post-remedial conditions
as well).

SITE SUMMARIES
  Results of applying the RAPS methodology to nine U.S. EPA
CERCLA sites are presented in Table 1. Although a contaminant
may travel through several environmental media before reaching
a sensitive receptor, the HPIs in Table 1 are associated with the
last medium through which sensitive receptors are exposed. For
example, a contaminant leaches from a landfill, migrates through
the groundwater and enters a surface water environment where a
local population is exposed. Although great environmental damage
may have occurred in the groundwater, the HPI for the analysis
is presented under the surface water column (Table 1).
  Table 1 also includes the population within a 50-mile radius of
each site. Although total population is presented part of the popu-
                                                                                    RAD AND MIXED WASTES    409

-------
lation may be exposed to any particular pathway. Note that the
overland pathway is not included in Table 1, but a column for local
soil  is  included. In  RAPS, the overland  transport pathway
represents an interim medium through which contaminants move.
This transport pathway only represents boundary conditions for
other transporting  media. Ingestion  of local  soils represents
exposure of a local population  residing on a contaminated site.
Table 1
HPI by Pathways
OrwiK!- Surflci
S.U Tolil HPI ConlMininl ปilซr lit.r
ABEX




Buflkir Hill


HutfMti Hiiir
Jotuiwn Control.
UliiulrMl
O.ii.il ftm
Church *Kk




D..I. Ll*..*








US Scriป







t7 4
11 t
11 f
2f 1
21 1
ป 9
U 9
4t (
U 4
-4 2
U i
n 9
17 7
17 t
11.7
1 ป
t t
U.7
41 7
27 t
ป 7
21 7
23 t
ป t
It 4
I t
tl 1
U 2
41 4
41 1
tl 9
U 7
24 1
17 2
ArMnic
ChrMiui
Liu>
Nickll
Z.ซ
CM!ซIU>
l.id
Zinc
PC a.
ThllliM
Trich limlhylMM
1,3,7.1-TCW) (Di.iin)
Ur.iMu.-IM
iMiw-m
Cntaiui
L.rt
S.l.n,ซ
ArMnic
1.1 OichlirMlhyllltl
Chl.rofon
v.njl ehUridi
BlfllCM
Tilrichlero.lhyl.M
TrichleroflArliM
Tin
Chronul
Birrlliu.
C.MlU.
U.4
Thill iu.
Antiuปr
ZlซC
Cw-r
SiUir
S.CI>
5.C
s.c
s,c
s.c
f,C
ซ,c
D.C
K.C
ป,c
u t
ซ,c
17 4
...

1
t
U
U
17.
...
21 t
...
It 1
...
...
,|C
1C
•.C
ซ,c
ซ,c
n.c
ซ,c
ซ.c
It t
...ซ
M.C
f.c
F.C
29 7
21 t
17 1
17 1
-4 2
17 1
II 1
2| 1
17 1
It 7
...
1 t

11
19
21
1
21
1
11
...
,C
.C
.c
,c
,c
.c
.c
.c
W-Ui
Air Locil tail Pwvllliwi
31
11
14
M
9 t
19
14
41
79
I.C
11 7
F,C
•.C
K.C
",c
K.C
ซ,c
c
n i
-l.i

...
i t
11 1
...
1 9
tl
U
41
41
41
It
24
17
17 4 1 1 i It*
11 1
It 9

19 7
17 1 1 1 i 11'
U 9
II 1
II 11.1*
I 1 1 . lป
1 1.1 1 11*
19 9 1 4 i 11*
H.C 1 1 > 11*
•.C
ซ,c
n.C
ซ.c
>,C III ll>
>.c
".C
ป.c
n.c
" c
>.c
•.c

I.C 7 7 • If
".C
ซ.c
l.C
".c
l.C
I.C
I.C
(.} S
(b) C


(c) -
(d) f
(•) •
     Ccutlrikvtiซft U hwain • ip
     •ccurrMl
     CMlM.liMIt ••• Ml Mdr..M4 by ih.t Hrlicwl.r
     f.thn/ M4ซl*ซt, fin.l lr.n.perlmg p.lhvif
     P.thvly Ml .odซIW
                                       ironปซfiUI t
-------
station at the valley floor near Kellogg, Idaho.
  Of three main heavy metals (lead, cadmium and zinc) at the site,
cadmium provided the highest HPI, primarily from inhalation of
wind-blown contamination from the exposed areas. Lead exposure
from soil ingestion by residents  of Pinehurst, Smelterville and
Kellogg was comparable to inhalation exposure. For cadmium and
zinc, contribution from soil ingestion was minor. A minor contri-
bution also resulted from surface water pathways (i.e., sport
fishing).

Church Rock
  The Church Rock Uranium Mill Tailings site is in northwestern
New Mexico near the town of Gallup. The mine and tailings are
managed by United Nuclear Corporation. The acid leaching process
used to extract uranium ore produces wastes with low pH, radio-
nuclides and heavy metals. The HRS ranking for this site is 30 over-
all. The groundwater, surface water and airborne scores were 43,
11 and 28, respectively.
  The groundwater was assessed by considering the potential risk
of using the deep aquifer beneath the site for domestic water supply
using measured contaminant concentrations found in monitoring
wells.  Surface water contamination  was evaluated using the
measured  concentrations found  in  the  discharge  from  mine
dewatering operations. Because no discharge flow rate was avail-
able and because the stream water is largely mine discharge, the
stream was considered to carry the same concentrations as the mine
waste outfall.
  Although documentation on the extent and condition of exposed
tailings along  the stream was inadequate to assess this pathway
the suspension of tailings materials along the stream banks could
be a significant pathway. Airborne measurements of total partic-
ulate loading  were available during remediation actions when
tailings along the stream were removed. In terms of the waste ponds
themselves, no significant atmospheric pathways were identified;
the potential release of radon gas from the pond tailings is con-
trolled by the water cover.
  The pathways considered  for this  site are groundwater usage
within 3  miles  and use of the Rio Puerco for recreation and sport
fishing. For most contaminants, the groundwater pathway provided
the major  contribution to the HPI. The groundwater analysis
assessed contaminants exposure  of 19 people  through drinking
water' while the surface water assessment addressed contaminant
exposure of 100 people through recreational and fishing activities.

Davis Liquid
  The Davis Liquid Chemical Disposal site occupies 15 acres in
a rural' residential section of Smithfield, Rhode Island. Liquid and
solid hazardous wastes were disposed into unlined pits from the
early 1970s to  December  1978, resulting in groundwater and sur-
face water  contamination at and surrounding the site. The wastes
are comprised of both chemical and sewage materials, including
paint and metal sludges, oils, solvents and liquid chemicals (pesti-
cides, acids, phenols, etc.). Contaminants have been released into
the environment from the waste unit and have  contaminated the
atmosphere (i.e. volatilization), nearby private wells and nearby
Latham  Creek; therefore groundwater surface water and atmos-
pheric pathways were addressed  using RAPS.
  For groundwater, measured contaminant concentrations in near-
by private wells were used to assess risk to those using the aquifer
under the waste unit for their domestic water supply. Both surface
water and groundwater feed Latham Creek, which flows 1.5 miles
to the Stillwater reservoir. Actual contaminant concentrations were
available in the creek and were used in the risk assessment.
  A potential atmospheric pathway exists as the result of volatili-
zation of organics from the waste storage unit.  Contaminant
volatilization rates were computed based on measured contaminant
concentrations in the waste unit soils; the RAPS atmospheric path-
way component was used to estimate exposure to the surrounding
population. Mechanical suspension of contaminants on a dirt road
running through the area also was addressed. Wind suspension of
contaminants (excluding mechanical suspension) on and around
the site is not expected because the soil in this swampy area is
generally wet.
  The most important exposure scenarios in calculating the HPIs
included contaminated groundwater from private wells, ingestion
of fish  from Latham Creek and inhalation of volatile organic con-
taminants by the population within 50 miles. The surface  water
and groundwater pathway analyses were based on measured water
concentrations and included exposure of only a few people. No
allowance was made for selected use of bottled water by the nearby
residents.  Relative contributions  to HPI by pathway  varied
depending on the contaminant of interest.

Hudson River
  From 1947 to 1977, two General Electric capacitor manufac-
turing  plants  discharged up to 1.1 million Ibs. of PCBs directly
into the upper Hudson River. Most PCBs were initially trapped
behind  a  100-year-old dam.  In  1973,  however, the dam was
removed, contaminated water and sediments were washed  down
the river and  PCB contamination became widespread. Contami-
nation  occurred in remnant sediments from the old dam, sediments
and dredge spoils along the river, the in-stream riverine environ-
ment and water carried into the Atlantic Ocean. The original U.S.
EPA HRS scores at the site were 76.92 and 55.00 for surface water
and atmospheric pathways, respectively.
  Current conditions were used to evaluate the health threat to
the surrounding population. PCB-contaminated river sediments
dredged from the Hudson River to remove PCB hotspots were not
included in the site evaluation, nor were any PCBs that washed
out the mouth of the Hudson River and into the Atlantic Ocean.
Transport pathways addressed by  RAPS included surface  water
and atmosphere. Because PCBs were discharged directly into the
Hudson River and monitored over the years neither the overland
nor groundwater environment required evaluation.
  The  site represents a 180-mile reach of the main portion of the
Hudson River in New York  State. The surface water portion is
divided into two designated  reaches: the upper Hudson River, which
is 30 miles long, and the lower Hudson River, which is 150 miles
long. Each reach has been contaminated by PCBs, the lower reach
to a lesser extent. In-stream sediment and water concentrations
measured at several locations along the river were used to calcu-
late health effects.
  This facility has two sites of concern for the atmospheric path-
way. The first site is a 40-mile stretch of the upper Hudson River
where  no atmospheric suspension of contaminated particles was
expected. The site documentation did, however, provide estimates
of PCB emission rates from the river; these volatilization rates were
used to compute population exposure.
  The  second site consists of five areas exposed when the dam
downstream from the plant was removed. These areas are dry and
open to particle  suspension and  volatilization. Suspension of
particles was  moderate; the volatilization rates of the PCBs were
addressed as  an old spill. Mechanical disturbances were not con-
sidered.
  The  primary contaminant at this site was PCBs in both river sedi-
ments and remnant sediments from a drained reservoir area. The
river sediment contribution (based on measured water concentra-
tions in the  Hudson River) was the major source of exposure.
Ingestion of river water and of contaminated fish harvested com-
mercially in the lower Hudson River were the most significant
exposure routes, although the atmospheric pathway (volatilization)
via inhalation and  farm product ingestion was  potentially
significant.

Johnson Control
  The Johnson Control site is an 18.5-acre industrial complex north
of Bennington, Vermont.  The surrounding area was exposed to
lead oxide emissions in 1977 and 1985. Liquid wastes of unknown
composition and amount also were discharged into a ditch behind
the plant from 1974 to 1977. Both surface water and groundwater

                       RAD AND MIXED WASTES     411

-------
pathways were considered exposure routes; however, the ground-
water pathway was discounted because the ditch is recharged by
groundwater flow, and contaminant migration would be confined
to surface water flow in the ditch.
  The only contaminant source information available for the ditch
was measured sediment concentrations, which were very low and
not necessarily representative of the original contaminated con-
centration present in the liquid waste discharge. Most likely, many
contaminants discharged to the ditch were carried downstream by
sediment transport  to Furnace Brook  and ultimately into the
Walloomsacre River. The RAPS surface water model does not
handle instream sedimentation at  this time so the surface water
pathway was evaluated by calculating the dissolved contaminant
concentrations  in the ditch and  their movement downstream
through Furnace Brook and the Walloomsacre River.
   Neither volatile nor semivolatile constituents were  reported at
this site, therefore eliminating volatilization as a potential release
mechanism to the atmospheric environment. Contaminated soil
suspension by wind also was discounted, because the ditch usually
contained water. There is no evidence of mechanical activity  or
physical disturbances in the area to  help induce resuspension  of
contaminated soil particles so the atmospheric pathway was not
modeled.
   The only release identified was contamination of Furnace Brook,
which flows  to the Walloomsacre River and then to the  Hoosic
River. The primary mode of exposure was from sport fishing on
the two rivers, neither of which is a source of drinking water.
Thallium was the most important contaminant.

Mainstreet Well Field
   The Mainstreet Well Field site in Elkhart, Indiana, includes 15
production wells for the city of Elkhart. Trichloroethylene (TCE)
contamination was initially discovered in 1981. This site is unique
in that the source of the contamination and extent of the plume
are poorly defined. There may be one or possibly two sources of
contamination. A TCE-contaminated industrial site east of the well
field seems to be the source; however the groundwater flow direc-
tion is south, southeast,  making contamination from the site
improbable. The original U.S. EPA HRS scores were 73.08 and
7.97  for  the groundwater  and  surface  water  environments,
respectively.
   Because Mainstreet is complex it was divided into two sites: the
well field and the industrial site east of the well field. For the well
field the groundwater, surface water, atmospheric and overland
pathways were not evaluated because HPI could be  determined
using measured  TCE concentrations in the wells.
   Because of extensive soil contamination, the industrial site could
affect groundwater, overland and atmospheric transport pathways.
Overland runoff to nearby Christiana Creek was  eliminated as a
pathway when aerial photographs showed that an elevated rail-
road bed between the site and the creek eliminated direct  runoff.
   Because the local hydrology is complex and not fully  understood,
the groundwater  scenario was addressed using conservative assump-
tions  to evaluate TCE migration from the surface soil into and
through the groundwater environment, eventually discharging into
and through Christiana Creek. Apparently the direction of ground-
water flow into the creek seasonally reverses depending on the
height of the creek. For the worst case, it was assumed that ground-
water continually recharges the creek. Principal contamination to
the atmospheric pathway is suspension of contaminated soil
particles and volatilization from old spills at the industrial area
(the well field is not considered in the atmospheric pathway). Sus-
pension is caused by wind erosion and mechanical suspension, the
latter from a paved street crossing contaminated soil.
   Volatile contaminants were reported at the site, and volatiliza-
tion calculations were performed. The high volatilization rates
computed reflected the type and quantities of contaminants at the
site and large amounts of contamination potentially saturated the
soil.
  Trichloroethylene was the major contaminant identified. The
primary exposure route was from contaminated well water. The
airborne and surface water (via groundwater) pathways were not
significant exposure routes.

Quail Run
  The Quail Run site  is a 27-acre former mobile home park 35
miles southwest of St. Louis, Missouri. In April 1971, its main road
was sprayed with 2,3,7,8-TCDD, contaminated waste oil to reduce
dust from mechanical  resuspension. In 1974 the road was paved,
and part of the road material was excavated and dumped nearby.
Site evaluation assumed that the original  population of  100
remained on-site and that the main road was not paved. Principal
areas  of contamination are  the unpaved road and its immediate
surroundings. The original HRS scores were 11.30, 8.39, 33.85 and
25.00 for groundwater,  surface water, atmospheric and direct-
contact  components, respectively.
  Overland,  surface  water and atmospheric pathways were
evaluated using RAPS. The groundwater pathway was not evalu-
ated because 2,3,7,8-TCDD has limited mobility and there were
no receptor wells in the shallow aquifer. Overland and surface water
environments were addressed by simulating the overland movement
of contaminated soil particles and water (caused by snowmelt and
rainfall  runoff) to the Little Fox Creek and  the migration  of
2,3,7,8-TCDD through Little Fox Creek and the Meramec River.
  The atmospheric pathway was employed to assess suspension
from the unpaved road. Volatilization occurred at very slow rates.
Because the environmental measurements were made after the road
was paved, the contaminant concentrations associated with the
unpaved road were estimated from the measurements made on the
dirt removed  during the paving process.
  This site was analyzed as though the paving of the road through
the trailer park had not occurred and therefore does not represent
the current  situation.  The  primary  mode  of exposure  to
2,3,7,8-TCDD is through soil ingestion by the trailer park residents.
Some exposure is also received from overland transport of con-
taminated sediment and water to the Little Fox Creek, which flows
into the Meramec River.  Drinking water, sport fishing and recrea-
tional contact were included in the overland surface water exposure
analysis. Contaminated  particle  inhalation via the atmospheric
pathway did not contribute significantly to exposures.
U.S.  Scrap
  U.S.  Scrap was an illegally operated industrial disposal opera-
tion located on approximately 9 acres in south Chicago Illinois.
The site was operated  from the late 1960s to 1980.  Paint industry
waste and pesticides were dumped into open pits or buried on-site.
The site received an HRS score of 1.92 in 1982. The low score is
attributed to the absence of groundwater and  surface water use
in the area. Because there  were no identifiable groundwater or
surface  water pathways for human exposure attributable to the site,
no  modeling of these pathways was required.
  The atmospheric component addressed the volatilization poten-
tial of organic contaminants from the waste site and suspension
of contaminated particles by wind. These emissions were based on
measured contaminant concentrations in the soil. For mechanical
suspension of contaminants, entrainment of contaminated particles
is not expected because there is  no vehicular traffic at the site.
  Airborne pathways evaluated for this site included organic com-
pound volatilization and trace metal particulate  suspension. These
computations are based on measured surface soil and subsurface
concentrations. The HPI values reported in Table 1 are all related
to particulate contaminants; the volatile organic contaminants did
not contribute significantly to exposure. Inhalation and, to a much
lesser degree,  farm  product  ingestion represented the  major
exposure routes for computing the HPIs.

CONCLUSION
  The  RAPS methodology was developed  to provide a better
management  tool  for  prioritizing  inactive  hazardous  and
412     RAD AND MIXED WASTES

-------
radioactive-mixed waste sites according to their relative risks to
surrounding populations and funding allocations for further site
investigations and possible remediation. RAPS addresses many of
the typical  limitations  associated  with  other simple ranking
systems3 as it  considers:

• More site information and constituent characteristics associated
  with potential contaminant transport pathways
• Chemical and radioactive wastes
• Potential direction of contaminant movement
• Contaminant mobility, persistence and toxicity
• Time until a population is exposed (i.e., contaminant arrival
  time)
• Exposure  duration

  This  paper  illustrates the application  of RAPS at several
hazardous waste sites; its flexibility in handling a wide variety of
complex problems, including geometric, geologic, hydrologic and
contaminant; and its application based on available data. Finally,
RAPS provides a structure or framework on which further inves-
tigations at a site can be based; that is  cause-and-effect relation-
ships can be developed. RAPS can then be used to help  focus
assessment exercises and help indicate  where potential problems
exist, why they are occurring, what effects changes  can have on
the assessment where to focus valuable resources (i.e., time and
money) and what alternatives may be most effective. This relatively
simple, but sophisticated, methodology provides a basis for com-
parison throughout the assessment process from beginning to end.
RAPS is not intended to be used only once, but it is intended as
a consistent approach for assessing the changes at a site and from
site to site, as assessment proceeds from beginning to end.


ACKNOWLEDGMENTS
  The authors thank K. Samec and R. Aiken for their guidance
and support in the development and testing of the RAPS metho-
dology and J. Gephart and S. Kreml for editing the  document.
This work is supported by the Office of Environment,  Safety and
Health,  U.S.  Department   of Energy under  contract DE-
AC06-76RLO 1830.


REFERENCES
1. Whelan, G., Strenge, D. L., Droppo J. G. Steelman, B. L. and Buck,
   J. W., "The Remedial Action Priority SYstem (RAPS): Mathematical
   Formulations," prepared for  the Office of Environment, Safety and
   Health, U.S. Department of Energy by Pacific Northwest Laboratory,
   Richland Washington. PNL-6200 1987.
2. Whelan, G., Steelman, B. L., Strenge, D. L. and Droppo, J. G., "Over-
   view of the Remedial Action Priority System (RAPS)"' in: Pollutants
   in a Multimedia Environment,. Y.  Cohen, (Ed.), Plenum Press, New
   York, NY, 1986,  191-227.
3. Whelan,  G.,  Strenge, D. L., Steelman B. L.  and Hawley. K. H.
' 'Develop-
   ment of  the Remedial Action Priority System: An Improved Risk
   Assessment Tool  for Prioritizing Hazardous and Radioactive-Mixed
   Waste Disposal Sites," In: Proc. of the Sixth National Conference and
   Exhibition on the Management of Uncontrolled Hazardous Waste Sites,
   Washington, D.C., November 1985, 432-437.
                                                                                           RAD AND MIXED WASTES     413

-------
          Role  of  Sedimentary  Channel  Deposite  in  Contaminant
                            Migration  and  Remediation  Design

                                               Marilyn A. Plitnik
                                    U.S.  Environmental Protection Agency
                            Region VI,  Hazardous Waste Management  Division
                                                  Dallas, Texas
                                                 Ramesh J. Shah
                                                 IT Corporation
                                                Irvine, California
ABSTRACT
  A RI/FS performed at a hazardous waste site in the Coastal Plain
of Delaware shows the site is to be located in an area that is very
conducive to contaminant migration. A Pleistocene channel deposit
lies directly beneath the site, and contamination has migrated into
this deposit as well as the underlying Low er Aquifers. The con-
taminant migration however, was not governed by the configura-
tion of the channel but was controlled by the potentiometric head
of the water table in the channel. The  water table contours are
parallel to the axis of the channel; therefore, the groundwater flow
is perpendicular to the long axis  of the channel.
  Where the channel deposits have truncated the confining clays,
the channel lies in direct hydraulic communication with the Lower
Aquifer and is a recharge area for underlying aquifers. Both the
Water Table Aquifer and the Lower Aquifer are contaminated
beyond the suspected disposal area. Thus, the paleochannei has
played an important role in the migration of the contaminants.
Similarly, the existence of the channel can be used beneficially to
design  a remedial measure. The contaminated  water can  be
extracted downgradient of the paleochannei. The treated ground-
water could be recharged on the upgradient side in the area of the
channel deposits. This design should retard further migration of
contaminants into the Lower Aquifer as the vertical head difference
in the contaminated area will be reversed or decreased because of
pumping by extraction wells.
  There is a misconception that groundwater will  always flow in
the direction of a channel deposit. In this case, groundwater flow
follows a fundamental groundwater flow principle and flows in
the direction of lower head. In this instance, the flow direction
is across the channel.  However, the presence of the paleochannei
at the site does control contaminant migration into the Lower
Aquifer. This same channel can be used to enhance (advance) the
contaminant removal from the aquifer.

INTRODUCTION
  The U.S. EPA conducted a RI/FS at a  Superfund site in northern
Delaware (Figure I) between 1983 and 1985 to establish the extent
of contamination  and to identify potential methods of remedia-
tion. The hazardous waste facility is located in a rural area.  In this
area the domestic water supply is obtained from the shallow water-
table aquifer, and high production wells tap the deeper semi-
confined to confined aquifers. In the initial stage, the preliminary
investigation by the State determined that the groundwater flow
direction in the shallow water-table aquifer was away from the near-
by residential development but toward the surface water creeks near
the site. To determine the extent of the contamination at the site

414    WATERWAYS AND WETLANDS RECLAMATION
and its potential impact on the deeper water supply wells, NUS
Corporation (NUS) performed a detailed remedial investigation
performed under a U.S.  EPA Superfund contract.
                         Figure 1
                    Site Location Map
  This paper evaluates the influence of the shallow paleochannei
deposits on the contaminant migration and on the remediation
measure developed. The outline of the channel is not controlled
by topography but is controlled by fluvial depositional environ-
ments. The contaminants follow a direction perpendicular to the
long axis of the channel. Hence, the contaminant migration is not
governed by the channel configuration but is controlled by the
direction of the groundwater flow  across the channel.  The
paleochannei, which contains coarse sand and  gravel, facilitates
recharge to the aquifers.
  This paper presents the concept of the paleochannei as a recharge
zone for treated groundwater. In addition, this paper demonstrates
how an  understanding of environmental deposition can be  used
to remediate a hazardous waste site.

-------
SITE DESCRIPTION
  The site is located in New Castle County, Delaware (Figure 1)
approximately 5 miles northwest of the Town of Kirkwood. The
site  is  also  situated approximately 0.5  mile east  of  the
Maryland/Delaware border.
  The site, depicted in Figure 2, is a plot of land within an iso-
lated area previously used for farming. However, the site is now
adjacent to a recent residential development called "Shelly Farms."
Wetland areas created by beaver dams exist to the west and south
of the site.
                                    A	A' OEOlOaC CHOS3-SECTION A-A

                                         ENCE
                          200     400    600 FEET
                          Figure 2
      Monitoring Wells and Geologic Cross-Sections Locations
  Site disposal operations were conducted between 1960 and 1977
in an open field. The facility accepted sanitary, municipal and in-
dustrial wastes as well as drummed wastes. Until 1963, combus-
tible wastes were burned as appropriate. Burning ceased after New
Castle  County revoked the permit to burn trash due to noxious
fumes. As a result of these operations, various stockpiles of waste
and debris were left on-site. Several hundred drums varying from
empty to full and in various stages of deterioration were also left
on-site.

REGIONAL GEOLOGY/HYDROLOGY
  The site is located in the Atlantic Coastal Plain Physiographic
Province. The Coastal Plain sediments, which unconformably over-
lie the crystalline basement, form a southeasterly thickening wedge
of unconsolidated sediments deposited along the Atlantic Continen-
tal Shelf. These sediments consist of unconsolidated clays, silts,
sands and gravels. Figure 3 shows a geologic cross-section of New
Castle  County, Delaware.
  The  oldest Coastal Plain sediments are Cretaceous deposits of
the Potomac Formation.  Sediments in this formation  consist
primarily of variegated silt and clay deposits with interbedded sand
strata,  which in many  locations are developed as an important
groundwater supply for New Castle County. Since the vertical and
horizontal distribution of the sands within the formation is highly
variable, the locations of these aquifers are not well defined or easily
predictable. In most cases, the sands are confined by a clay layer
(the Potomac clay) and behave like an artesian aquifer. Generally,
the Potomac clay is a distinctive grey or red color, highly plastic
and stiff.
  Pleistocene-age Columbia deposits locally overlie the Potomac
Formation.  These deposits consist of orange, tan and  yellow,
medium- to coarse -grained sands and gravels  that were deposited
by streams which formed a system of straight channels. The contact
between the Columbia deposits and the Potomac Formation is an
erosional unconformity, hence the thickness of the Columbia
deposits is very irregular. Sands are generally thickest in the buried
channels. These channels contain coarse-grained sands, gravels and
pebbles. The Columbia deposits form an unconfined or water table
aquifer. The groundwater-table surface generally follows the
ground surface. Most of the groundwater is recharged by natural
                                                                      APPROXIMATE LOCATION OF SITE


                                                                      (LATERALLY PROJECTED ABOUT SIX MILES)
                                                                                             MIOCENE CHESAPEAKE 4NOUP
                                                                                          .PLEISTOCENE (COLUMBIA) SEDIMENTS
                                      SOURCE: BJNDSmOM AMD PflCKETT. 1871
                                                                                                                          LEVEL
   •nor
                                                             Figure 3
                                         Geologic Cross Section of New  Castle County, Delaware
                                                                     WATERWAYS AND WETLANDS RECLAMATION     415

-------
    taCA LEVEL 0-
                           - "**
                 mam LfVIU UlAJUMD ON 4-1-11
                                                                                *"•-*••
                                                                                         I ซATn€ปEO
                                                                                         I BEOKOCK
 KEY TO WTTERNS: [    ] SAND     |    | CLAY     |=||| SHOT OJY tSfel] 5I.TY SAM)  |    |


KEY TO FORMATIONS' PC • COUUMBW fOBMATON   KptWPOTDMAC FORMATION (U*>O> POTOtttC DEWSTTS)   KtdJ-WIDMAC FOWATIOK(UJWCT COIซ*ซC OtKarS          M
                                                                       Figure 4
                                                             Geologic Cross-Section A-A'
            ) SEA ICVCL 0-


 KEY TO PATTERNS'  [    |sAND


KEY TO FORMATIONS' PC • COLUMBIA FORMATION   KplUI-POTOMAC FCMMATON (UPPER POTOMAC OEPOSTTSI   KplLl'POTOMAC PORMATIONIuOWEII POTOMAC DEPOSITS)    $Q™
                                                                                                                             NONItONTIL SCALC M
                                                                       Figure J
                                                             Geologic Cross-Section B-B'
                                                                                                                         Vf DTICU tCAU IN rtCT
  416     WATERWAYS AND WETLANDS RECLAMATION

-------
precipitation.  The Columbia  aquifer supplies water to most
domestic wells and contributes to the base flow of area streams.
The Columbia and Upper Potomac aquifers constitute a potable
water source for residents in the immediate area, and the Lower
Potomac aquifer is a production zone for several large supply wells
near the site.

REMEDIAL INVESTIGATION SUMMARY
  The intent of the RI was to determine the nature and extent of
contamination at the site to evaluate appropriate remedial actions
to minimize health and environmental risks. The RI field activities
consisted of: collecting surface soil samples, collecting surface water
samples, collecting sediment  samples,  collecting groundwater
samples, installing additional monitor wells, conducting pumping
tests, performing geophysical  surveys excavating test pits and
collecting subsurface soil samples.

Site  Geology
  Data collected from drill cuttings, Shelby tube sampling, split-
spoon sampling, surficial geophysical studies and borehole geophys-
ical logging were used to characterize the site geology. The geology
is shown on cross-sections A-A' (Figure 4) and B-B (Figure 5)
  The Potomac Formation underlying the site can be divided into
a Lower Hydrologic Zone  (LHZ) and Upper  Hydrologic Zone
(UHZ). Deposits in the LHZ are grey to dark grey, fine to medium
sand, lignitic silty sand and lignite interbedded with stiff grey clays
and silty clay. A silty clay occurs as the  basal confining layer of
this  unit. This unit is approximately 45  ft thick at the site.
  The Potomac clay, the major confining layer between the UHZ
and the LHZ,  is a very  stiff, variegated grey,  brown and red silty
clay or clay with discontinuous  lenses of fine to medium sands and
silty sands. Two distinct sand  lenses of lateral significance were
encountered in this clay. The uppermost lens occurred just beneath
the surface of the confining clay and is seen on cross-section B-B'
(Figure 5). The second lens was encountered in monitor wells 1O1D,
 107D, 108D and 109D.
  The Upper Potomac deposits subcrop  north of the fenced area
as can be seen in Figure 4. These deposits occur downdip to a thick-
ness of 60 ft, south of the fenced area, near well 108D. The sand
and silt deposits of the Upper Potomac have been separated from
the Columbia  deposits  by a clay lens; however, the lens pinches
out  west and north of  well  108D.
  The uppermost formation occurring at the  site is the Columbia
deposits. Figure 6 shows the contour map of the base of the Colum-
bia deposits. These deposits consist of brown to orange, very dense,
fine to  coarse grained sands, gravels, silty sands and occasional
cobbles. The Columbia sediments in the site  area were deposited
in a fluvial channel, trending east-west. The channel does not
coincide with a topographic low and is therefore a result of the
depositional environment. The  channel material consists of coarse
sand, gravel, and cobbles. This  channel has truncated the Potomac
Formation by removing the upper confining Potomac clay deposits
and replacing them with stream sands and gravels. These relatively
flat-lying stream deposits  unconformably  overlie the dipping
Potomac Formation. The Columbia deposits  under the site range
in thickness from 19 feet to 46 ft. Cross section A—A', a geologic
cross-section across the channel deposit,  shows the varying thick-
ness of Columbia deposits. Cross section B—B' is a geologic cross-
section parallel to the channel deposits and, therefore, shows almost
uniform thickness of Columbia deposits. Samples collected at the
base of the channel that showed reworking of the Columbia sands
and gravels with Potomac Formation deposits are evidence of the
truncation.
Site Hydrogeology
  Based on the correlation of  subsurface geology and hydraulic
connection of the Upper Potomac sands with  the Columbia sands
and gravels, these deposits constitute a single hydrogeologic zone.
However, the coarser sands and gravels of the Columbia deposits
have a higher permeability than the fine- to medium grained sands
                  -i*Vซ'iป/N '
               •3^1^ jfc
               -----A  ' •  ^rat*,'.j*~ .	
                               **-.W-  TIWTOซO
                                     MOMTOHN3WELL

                                — 35 — CONTOUR S40WIN3 ELEVATIONfMSU FEET

                                     COLUMBIA CHANNEL
                            Figure 6
             Base Contours Map of Columbia Deposits

of the Upper Potomac; therefore, separate potentiometric surface
maps were prepared.
  The groundwater flow direction in the Columbia deposits at the
site is to the south and southeast toward the wetland areas and
Long Creek (Figure 7) and southwest toward the beaver pond. The
change  in water level elevations across  the site  is primarily a
response to changes in topographic elevations. A steeper hydraulic
gradient in the northern portion of the site exhibits this trend. In
other portions of the site, the horizontal hydraulic gradient is less,
due to smaller topographic changes and proximity to creeks which
are groundwater discharge zones.
                          100     400    (00 FEET

                           Figure 7
          Potentiometric Surface Map of Columbia Aquifer
                                                                    WATERWAYS AND WETLANDS RECLAMATION     417

-------
  Groundwater flow directions in the Upper Potomac deposits are
similar to those of the Columbia deposits but are less influenced
by topography (Figure 8). In most of the study area, the entire
Upper Potomac Formation is  under water table conditions.
However, southwest of the site, a clay lens separates the Columbia
deposits from the Upper Potomac sands. Therefore, in this area,
the Potomac sands exhibit semi-confined behavior.
                          Figure 8
      Potentiometric Surface Map of Upper Potomac Aquifer
  Vertical groundwater  movement between the Columbia  and
Upper Potomac Formations was assessed by differences in water
elevations from wells constructed in both formations at the same
drill site. Where the water elevation is lower in wells screened in
the UHZ sands versus those screened in the Columbia deposits,
a vertical downward gradient exists. Where the water elevation is
higher in the UHZ versus the Columbia deposits, an upward vertical
gradient exists. Downward gradients indicate a potential recharge
zone and upward gradients indicate a potential discharge zone.
Based on water level data, a downward vertical gradient (poten-
tial recharge zone) exists across the site in the area of the channel
deposits.
  Both  these aquifers (Columbia and UHZ), which are hydrauli-
cally connected for the most part in the study area, discharge into
nearby streams, thus supporting base flow. The surface water dis-
charge zones include the wetlands, the beaver ponds and Long
Creek.
  Precipitation is the source of recharge to the Columbia aquifer,
which in turn recharges the Potomac aquifers. Annual precipitation
averages 44.5 in. of which approximately 14 in.  are available for
recharge.
  Groundwater seepage velocities in the Columbia aquifer range
from 550 ft/yr southwest of the site to 2,300 ft/yr to the southeast
of the site.  Groundwater seepage velocity  in the UHZ aquifer
averages about 240  ft/yr.
  The volumetric flow rate of groundwater through the Columbia
deposits in the channel beneath the site was calculated to be 68
thousand gal/day or 24 million gal/year. The volumetric flow rate
of groundwater through the Upper Potomac deposits was calcu-
lated to be approximately 11  thousand gal/day  or  4 million
gal/year. These data indicate the predominant influence of the
channel deposits on the flow regime at the site area.
  Sand and silty sand deposits in the LHZ are under confined con-
ditions.  Water levels show a downward vertical gradient from the
UHZ.  However, groundwater movement  through  the  clay
separating the two zones is very slow. It is estimated,  based on
available permeability and vertical-gradient data, that groundwater
movement in the LHZ is to the northeast at approximately 130
ft/year.

Groundwater Contamination
  The problem, as identified by the RI, is that contamination of
soil, surface water and groundwater has occurred as a result of
improper disposal of wastes at this site. The contaminants include
organics and inorganic compounds  related to paint pigments,
sludges  and solvents.
  Chemical contaminants identified in the groundwater correlate
with the contaminants found in the surface and subsurface soils.
Toluene, ethylbenzene, xylenes and methylene  chloride were
detected in  the groundwater  in high concentrations.  Benzene,
phthalate esters and other organic compounds were detected at
lower concentrations. Trace metals, lead and chromium also were
detected in the groundwater.
  Contamination of groundwater has occurred in both the Colum-
bia and UHZ aquifers. As discussed in the hydrogeology section,
the channel in the Columbia deposits across the site has played
a key role in the migration of contaminants from the site. Since
a vertical downward gradient exists between these two aquifers and
the LHZ, the currently unconlaminated LHZ also is threatened.
The existence of the channel facilitates the movement of con-
taminants into the Columbia  and UHZ aquifers. This area acts
as an ideal recharge zone due to the coarse material it contains
and its location. Contrary to popular  belief that contaminants will
flow down the channel, once the contaminants are in the channel,
they flow in the direction of the  water table gradient, that is,
perpendicular to the long axis of the channel and not down the
channel itself.

DESIGN OF GROUNDWATER PUMPING SYSTEM
  Based on the findings of the RI, a ROD signed by the U.S. EPA
in September 1985 has selected a groundwater recovery system as
a part of the remediation strategy. A recovery well system was
selected to extracting contaminated groundwater from beneath the
site. The actual design specifications of the system will be developed
during  the design phase; however, the conceptual design of such
a system is presented here. The existence of the channel can be
used beneficially in the design of this contaminant extraction
system.  The overall  objectives of this extraction well system are
to capture and remove the contaminants, lower the groundwater
table to reduce and eliminate discharge to the surface water bodies
and lower the vertical hydraulic gradient between the Columbia-
UHZ and the LHZ aquifers This system also will serve to flush
the contaminated soils. By recharging the aquifer with the treated
groundwater, using the channel deposits as a recharge area, the
flushing of the contaminants  from the soils will be accelerated.
  The proposed  extraction system consists of four wells, each 10
to 12 in. in diameter with an average depth of 40 ft and an average
screen length of 20 feet. Each well is designed to operate at an
optimal pumping rate of 350 gal/min. The  wells will be pumped
in pairs so that the maximum groundwater extraction rate at any
time will be 700  gal/min.  Figure 9 presents the proposed extrac-
tion well locations. Well E-l is located in the middle of the fenced
area where groundwater contamination is most severe. Well E-2
is located to the west between MW-2 and MW-6. Well E-3 is located
to the east  side of the fenced area between  MW-1 and MW-101.
Well E-4 is located south of the fenced area near MW-7. Wells
E-l and E-2 are located in areas of very high contamination. Wells
E-3 and E-4 are located just downgradient of the channel deposits
to capture the plume as it moves out of the channel in the down-
418    WATERWAYS AND WETLANDS RECLAMATION

-------
gradient direction.
  The determination of the numbers and locations of wells neces-
sary to fully capture the contaminated groundwater was facilitated
by the use of the drawdown curves.  Well locations were selected
such that the cones of depression significantly overlap in the areas
of greatest contamination.
  The extracted groundwater will be treated in an on-site water
treatment plant and will be recharged via a spray irrigation system
on a grossly contaminated area which lies directly above and slightly
upgradient of these Columbia channel deposits.
  Given the available data, a definitive prediction of  the length
of time or  number of flushes necessary to restore the  aquifer to
                                     TREATMENT PLANT AREA

                                  E-l EXTRACTION WEU. (PROPOSED)
                            Figure 9
                Proposed Extraction Well Locations
acceptable levels is not possible. However, the limited extent of
the plume, the characteristics of the contaminants, the characteristic
of the soils and the hydrogeologic conditions at the site are such
that this system should prove effective in a relatively short period
of time.

CONCLUSIONS
  Sites in the Coastal Plain deposits pose a real threat to valuable
groundwater resources, especially  in areas where the population
is highly dependent on groundwater for its drinking water supply.
A thorough subsurface investigation is necessary to fully identify
existing geologic and hydrogeologic conditions prior to proposing
any remedial actions.
  Information gained through investigations at this Superfund site
revealed that contrary to the common belief that groundwater will
flow along the coarse paleochannel deposits (least resistance), the
groundwater in fact flowed perpendicular to the channel. By  un-
derstanding the existing hydrogeological conditions of the site and
the important role of the paleochannel deposits as a recharge area,
optimum use of these conditions could be made to enhance remedi-
ation of the groundwater and contaminated soils.

REFERENCES
1.  NUS Corp., "Remedial Investigation and Feasibility Study—Harvey
   Knott Drum Site, New  Castle County, Delaware." Volumes I-IV.; U.S.
   EPA contract Number 68-01-6699, 1985
2.  Shah, R. J., "Hydrogeological Investigation, Harvey-Knott Site, New
   Castle County., DE;" Unpublished Report; Water  Supply Branch,
   Delaware Department  of Natural Resources and Environmental Con-
   trol,  1982
3.  Spoljaric, N., "Pleistocene Channels of New Castle County, Delaware,"
   Geological Survey, Report of Investigation No. 10, 1967
4.  Spoljaric,  N. and Woodruff,  K. D., "Geology, Hydrology, and Geo-
   physics  of Columbia Sediments in the Middletown-Odessa Area, Dela-
   ware;"  Delaware Geological  Survey, Bulletin. 13, 1970
5.  Sundstrom, R. W. and Pickett, T.  E., "The Availability of Ground-
   water in New Castle County, Delaware;" Water Resources Center,
   University of Delaware, Newark, DE, 1971
                                                                     WATERWAYS AND WETLANDS RECLAMATION    419

-------
                              Application of a  Mixed-Method
                            Analytical  Scheme  for  Analysis of
                          PCB  in  Water and  Sediment  Samples
                                    From  a  Polluted Estuary

                                              Richard A. McGrath
                                             William G. Steinhauer
                                            Battelle Ocean Sciences
                                            Duxbury, Massachusetts
                                            Siegfried L.  Stockinger
                                         Ebasco  Services Incorporated
                                            Dedham, Massachusetts
 ABSTRACT
  As part of the Remedial Investigation/Feasibility Study for the
 New Bedford Harbor Superfund site, a field sampling and analyti-
 cal chemistry program was conducted to provide data for calibra-
 tion  and application of mathematical  models of contaminant
 transport and fate.  Gas chromatography with electron capture
 detection (GC-ECD) was originally selected for analysis of PCBs
 by congener group. After the samples had been collected, the U.S.
 EPA requested a change of method to recently developed U.S. EPA
 Method 680  using gas chromatography/mass  spectrometry
 (GC/MS). Application of Method 680 to approximately 700 field
 samples yielded data with method  detection limits too high for
 proper implementation of the models.
                   70'55
if id -
41'35
                       Figure 1
       New Bedford Harbor Superfund Project Study Area

420    WATERWAYS AND WETLANDS RECLAMATION
  After conducting a study to compare data obtained by the two
methods, an alternative analytical scheme was developed that relied
on both Method 680 and the original GC-ECD method. An itera-
tive screening procedure with a  series of volume-adjustment and
cleanup steps was developed in  conjunction with the application
of the GC-ECD  analysis. Data generated by the two methods,
indicating similarity of results,  are presented and compared.

INTRODUCTION
  The Acushnet  River Estuary, which in its middle and lower
reaches forms New Bedford Harbor, Massachusetts (Figure 1), is
heavily contaminated  with PCBs and several heavy metals. New
Bedford has been a major east coast fishing and shipping port since
colonial times and remains one of the leading fishing ports in the
United States. In addition, the harbor is surrounded by a variety
of light manufacturing and processing industries including textiles,
dyeing, electroplating, metal  finishing and electrical component
manufacture.
  Two electrical component manufacturers. Aerovox Incorporated
and Cornell-Dubilier  Electronics  Corporation (Figure  1),  used
PCBs in the manufacture of electrical capacitors for a period of
approximately 40 years ending in 1977 when the use of PCBs was
banned in the United  States,  For most of this period, these and
possibly other industries discharged waste water contaminated with
PCBs to the harbor and the  municipal sewerage system1.
  PCBs are a class of compounds produced by chlorination of the
biphenyl molecule. There are 209 distinct PCB isomers that vary
in the number and position of the chlorine atoms; the chemical
structure of the biphenyl molecule and some typical PCB isomers
are shown in Figure 2. PCBs were used commercially in the form
of isomer mixtures produced in the United States by Monsanto
Corporation under the trade name of Aroclor. Different Aroclor
mixtures are designated by four-digit numbers; the last two digits
indicate the approximate average percentage (by weight) of chlorine
in the blend. The sole exception is Aroclor 1016, which is approxi-
mately 41 % chlorine. In New Bedford, the manufacturers primarily
used Aroclor 1242, with lesser amounts of Aroclors 1252 and 1254,
until these mixtures were replaced with Aroclor 1016 in 1971.
  Elevated concentrations of PCBs, usually measured as Arodor
1254, were first reported in sediments from New Bedford harbor
in 19762. Since then, many investigations have documented the
widespread PCB contamination of water, sediments and biota in
the Acushnet River, New Bedford Harbor and adjacent areas  of
Buzzards Bay. PCB concentrations in sediments in the Acushnet
River frequently exceed 500 ppm  dry weight and occasional samples
exceed 100,000 ppm (10%)'. Of the local biota, eels (Anguilla

-------
CH2— CH
                                   CH., — CH2
                   CHj— CHj         CH2— CH2

                       Biphenyl Molecule
                   Monochlorobiphenyls (Cl,- PCB)

                       Cl               ci
                       Cl                     ci
                 Tetrachlorobiphenyls (CI4- PCB)
                 Heptachlorobiphenyls (CI7- PCB)

                          Figure 2
         Structure of Biphenyl Molecule and Examples of
               Polychlorinated Biphenyls (PCBs)
rostrata) appear to be the most heavily contaminated, with body
burdens occasionally in excess of 500 ppm; lobsters (Homarus
americanus) have been reported with concentrations in muscle tissue
exceeding 50 ppm1.  The FDA action level for PCBs in  edible
tissues of fish and shellfish is 2 ppm.  Because of the high levels
of PCB contamination, the New Bedford Harbor area was closed
to fishing for certain species in 1979.
  Because of the magnitude  of PCB contamination in the harbor
sediments and the difficulties inherent in dealing with such con-
tamination in a tidal estuary,  New Bedford Harbor currently is
one of the largest and most  complex Superfund projects. Ongo-
ing studies in support of the RI/FS leading to a ROD scheduled
for late 1988 involve numerous subcontractors and technical dis-
ciplines. A key component of the New Bedford Project is the de-
velopment      and      application       of      linked
hydrodynamic/sediment-transport and food chain models to  in-
vestigate the long-term fate of PCBs in the harbor under a no-action
baseline and two or  more remedial action alternatives.
  The modeling program consists of several distinct tasks. Field
studies were performed  from 1984 through 1986 to determine
normal and storm-induced current patterns in the harbor and bay.
A sampling program  was conducted during the same period to pro-
vide chemical data for input to the models. Laboratory experiments
were conducted to obtain site-specific information on adsorption/
desorption kinetics and bioaccumulation factors. An  analytical
chemistry program,  the subject of this paper, was undertaken to
analyze  the approximately 3000 samples generated in the field
program  and  a data management system  was designed and
implemented to allow dissemination and manipulation of the data.
These tasks have largely been completed. Remaining activities
include the final calibration and validation of the two models and
application of the models to the baseline and remedial action cases
at the site.
  Although much earlier work on PCBs has dealt with the vari-
ous isomers either as a single  contaminant ("total" PCB) or as
Aroclor mixtures, the disadvantages of this approach have been
recognized for some time. In particular, studies of the transport
and fate of PCBs in the environment are adversely affected by the
extreme variation in  the physical, chemical  and biochemical
behavior of the individual isomers3'4. It was apparent early in the
design  of the New Bedford  Harbor Modeling Program  that
modeling either total PCB or one or more Aroclors would require
assigning "average" physical, chemical properties to large groups
of compounds and would lead to less than satisfactory modeling
results. However, the cost of modeling individual isomer would
have been prohibitive, both  in  the modeling  and the analyses
required to support it. As a compromise, it was decided  to con-
sider PCBs as homologue or pseudocomponent groups according
to   their    level    of   chlorination    (e.g.,   C1,-PCB,
C12-PCB,...C110-PCB).
  Physical constants for solubility and partitioning of the 10 PCB
pseudocomponents (Table 1) demonstrate the validity  of  this
approach. Mean solubilities typically differ by a factor of 2 to 10
between adjacent pseudocomponent groups and by approximately
five orders of magnitude over all groups. The ratio of PCB sorbed
to organic carbon in sediments versus non-sorbed PCB (Koc) varies
by nearly seven orders of magnitude and the amount sorbed onto
colloidal organic matter varies by approximately two  orders of
magnitude. Because the less chlorinated PCBs (Cl, and C12) are
readily volatilized in the environment and the  more chlorinated
PCBs (>C16) are a relatively minor component of the commer-
cial Aroclors used at New Bedford, it was decided to model only
the C13 through C16 pseudocomponents. The analytical  program,
however, was designed to produce data for all 10 groups.
                                                                       Table 1
                                                     Estimate of the Degree of Sorption of PCB Isomers
                                                      Onto Colloidal Organic Matter of Concentrations
                                                                1.0 and 10.0 mg ORG C/l
No. of Chlorines
(hoaologues)
1
2
3
4
5
6
7
8
9
10
Solubility K Percent PCB Sorbed
(ppb) (Calculated) 1.0 ซg/l 10.0 *g/l
3000 4,
1000 1.
200 4.
50 1
10 5,
1.0 3.
0.5 5.
0.2 1.
0.1 2.
0.016 1.
Note: Solubility and K of individual
grouping varies
.77 x 10?
.17 x 10*
.40 x 10*
.37 x 10^
.13 x 10J!
.39 x 10*
.98 x 10:
.27 x 10'
,24 x 10'
,01 x 10
compounds
.475 4.55
1.16
4.21
12.0
33.9
77.2
85.7
92.7
96.5
99.0
vithin each
10.5
30.6
57.8
83.7
97.1
98.4
99.2
99.6
99.9
honologue
                                                Overall management of the New Bedford Harbor project is the
                                              responsibility of Ebasco Services Incorporated under their REM
                                              III contract with the U.S. EPA. Battelle Ocean Sciences coordinates
                                              and manages  the  modeling program  and conducted the field
                                              sampling, laboratory experiments, most analyses using U.S. EPA
                                              Method 680 and the comparability study reported herein. Battelle
                                              Pacific Northwest Laboratories and HydroQual, Inc., are respon-
                                              sible for the development and application of the physical trans-
                                              port model and food chain model, respectively. Aquatec, Inc.,
                                              conducted the GC-ECD analysis of New Bedford Harbor samples
                                              as part of the U.S. EPA Contract Laboratory Program.

                                              ANALYTICAL METHOD SELECTION
                                                The original PCB analysis method proposed for supporting the
                                              modeling effort was based on gas chromatography using electron
                                                                  WATERWAYS AND WETLANDS RECLAMATION    421

-------
capture detection (GC-ECD). The proposed GC-ECD method
provided for PCB level of chlorination determination in a man-
ner analogous to methods reported recently5'6. In conjunction
with the modeling teams at Battelle Pacific Northwest Laborato-
ries and HydroQual, Inc., it was determined that sample sizes of
0.05, 0.015 and 20.0 kg for sediment, tissue and  seawater, respec-
tively,  would yield detection limits (total PCB) for each sample
type on the order of 2.5 ppb, 15 ppb and 4 ppt,  respectively. The
field sampling program, comprising three major surveys of approx-
imately 1 month duration each, was developed based on these sam-
ple sizes.
   The first field survey of the harbor was conducted in Septem-
ber and October of 1984. At approximately the same time, U.S.
EPA requested that a method then being validated for the analysis
and quantification of PCB by level of chlorination,  U.S. EPA
Method 680, be considered for the analysis of samples from New
Bedford. Although the use of Method 680 was considered early
in the  modeling program, it was not until the  summer of 1985,
at about the time of completion of the final field survey, that direc-
tion was received to  proceed with these analyses by Method 680;
it was  not until the fall of 1985 that the method was validated in
the laboratory and the first field samples were analyzed.
   Because this change in method affected the ability of the
analytical laboratory to meet Contract Required Detection Limits
(CRDL), a request was made to U.S. EPA to change these detec-
tion limits to correspond to those that could reasonably be expected
using Method  680. These calculated detection limits are presented
in Table 2. Both  full  scan and  selected  ion monitoring (SIM)
CRDLs were calculated from laboratory calibration data. Based
on this information, U.S. EPA approved the use of Method 680
for the analysis  of all New Bedford field samples.
on an analysis of historical data, it was determined that PCB levels
in sediments from stations south of Clarks Point (Figure 1) would
allow quantification by Method 680 of some homologues, but not
the full complement of homologues selected for modeling. Further,
assessment of the historical data indicated that the variation in PCB
levels in  resident biota was too great to ensure that PCB levels
would be high enough for quantification using Method 680. Esti-
mated ranges of PCB concentrations in New Bedford Harbor sam-
ples as developed during this evaluation are presented in Table 3.
                           Table 3
   Ranges of PCB Concentrations in New Bedford Tissue Samples
            Arซa 1        Area 1       Area 3       Area 4
        (Inner Harbor)   (Outer Harbor)  (Barbor/Bay)  (Bunards Say)
Lobstar
Crab
Flounder
Clai
Hussel
Seavora
*Vet velghl
1-10
1-10
5-20
0.1-10
0.1-10
1-10
edible tissue Cor
0.6-20
0.5-10
0.2-8
0.001-5
0.001-5
0.5-10
lobster ,
and flounder approximately 1/3 conctn
0.1-10
0.1-1
0.1-10
0.001-1
0.001-1
0.1-10
0.02-7
0.05-0.5
0.01-20
0.02-0.05
0.02-0.1
0.01-10
crab, flounder. Vet velght vhole
PCB In vhole body for lobster, crab,
tratlon In edible tissue.
                           Table 2
        Contract Required Detection Limits for PCB Level of
        Chlorination Determined by Method 680. Values are
          Given for Full Scan Mode and, In Parentheses,
                  For Selection Ion Monitoring
Congener
Cl -Cl
C14-C16
Cl* Cl'
"Xo
Sediaent
(ng/g dry)
2(0.6)
*(1.3)
6(2.0)
10(3.3)
Tissue
(ng/g wet)
2.5(O.B)
5(1.7)
7.5(2.5)
12.5(4.2)
Seftvater
(ng/llter)
2.8(0.9)
5.6(1.9)
8.4(2.8)
14(4.7)
*CRDLs based on the following assumptions:
               1) Preinjection volume Is 100 ul for all samples;
               2) 2 ul Injected for all samples;
               3) Average sediment sample size is 50 g (SOX water);
               4) Average tissue sample size Is 20 g;
               5) Average aeavater sample size Is 18 lltcrsi
               6) Extraction efficiency Is SOX for all preparation
                 procedures.
IMPLICATIONS OF THE CHANGE TO METHOD 680
   Although prompted by the desire to provide the best possible
data set for multiple users, the change to Method 680 soon proved
to have profound implications for the modeling program. GC/MS
analysis of New Bedford/Buzzards Bay seawater filtrate and par-
ticulate samples using U.S. EPA Method 680 showed that PCB
concentrations in approximately  half  of the samples were the
method detection limit (MDL). In general, levels of PCB in both
paniculate and filtrate samples from New Bedford inner harbor
(i.e., inshore of the hurricane barrier) (Figure  1) were high enough
to be quantify by GC/MS. For stations in the outer harbor and
Buzzard's Bay, the specified sample sizes, designed for GC-ECD
analysis, resulted in concentrations in extract that were below MDL
for one or more of the homologue groups to be modeled. Based
  The net effect of the detection limit problem was that approxi-
mately one half of the data points for PCB in seawater needed
for the model would be reported only as being below MDL, which
was approximately 70 ppt total PCB. Collection of larger volume
seawater samples  would have required repeating the entire field
program, an option that was rejected due to cost and schedule con-
siderations. Use of the MDL or one-half the HDL as input to the
physical/chemical model was  considered a possible, though less
than ideal, solution. As part of the decision process, HydroQual
ran the preliminarily calibrated food-chain model using the HDL
as the ambient seawater concentration. The results of these tests
showed clearly that use of the seawater MDL would lead to con-
centrations of total PCB in biota far in excess of the amount indi-
cated by the data available at that time (results of analyses of biota
samples collected for the modeling program were not yet available).
  Two issues prompted U.S. EPA to rethink the strategy for sam-
ple analysis in the RI/FS study. First, additional information for
those samples below MDL already analyzed by Method 680 was
necessary for modeling. Second,  it was important to ensure that
level of chlorination information for all four homologue groups
to be considered in  the model could be obtained for all samples
collected in  the field program. Alternatives to Method 680 (full
scan), included Method 680 (SIM detection) and validated use of
GC-ECD or equivalent detection. Even the increased sensitivity
of SIM was not expected to be sufficient for the purposes of the
modeling program. As a result, a decision was reached that those
samples that had not been or could not be quantified by Method
680 would be analyzed by GC-ECD with its approximate sensitiv-
ity of 2 pg/1 for seawater. For verification of the GC-ECD method
and to compare results obtained  using the two methods, a com-
parability study was conducted using selected seawater samples with
a range of expected PCB concentrations suitable  for analysis by
both methods.
422    WATERWAYS AND WETLANDS RECLAMATION

-------
METHOD COMPARISON (COMPARABILITY STUDY
Overview
  The comparability study involved the analysis of 22 field sea-
water samples (12 dissolved phase and 10 suspended particulate
phase) and four blanks (two of each phase). The samples were
selected from stations in the inner harbor representing the range
of expected PCS concentrations that were sufficiently high to pro-
vide data from each method; two samples from the station with
the lowest expected concentration (Buzzards Bay boundary) were
analyzed also to provide some comparison of actual concentra-
tions versus MDLs for all homologues. Analysis of method blanks
by each method was intended to confirm that neither method intro-
duced contamination interfering with quantification of PCBs in
the samples.
  The GC-ECD method used in the study was developed originally
for the modeling program  and is essentially similar to that recom-
mended by the U.S. EPA for use in work supporting the 301(h)
program5. The quantification technique relies on high resolution
(capillary column) GC-ECD  and a determination of response
factors for resolvable PCB peaks. The resolved peaks in a PCB
calibration standard composed of Aroclors  1242, 1254 and 1260
were identified by GC/MS and quantified by level of chlorination
by Method 680. Serial dilutions of the same calibration standard
were then analyzed by GC-ECD. The GC/MS  results were used
to determine response factors (RFs) for PCB at each level of chlo-
rination. The field samples were then analyzed and quantified by
GC-ECD. PCB by level of chlorination was calculated as the sum
of all resolved, RF-corrected PCB  peaks at each level of chlori-
nation.

Method
   The calibration solution of Aroclors was  analyzed on a Finni-
gan 4530 gas  chromatograph/mass spectrometer that had been
tuned and calibrated according to methods specified in U.S. EPA
Method 680. The level of chlorination and concentration  of PCB
in each resolvable peak was then determined according to  method
protocols. GC-ECD standards were obtained by serial dilution of
the GC/MS Aroclor calibration standard.
   Instrument operating conditions for the analysis of PCBs by GC-
ECD are presented in Table 4. The Aroclor  standard was diluted
to make three calibration standards with concentrations within the
GC-ECD calibration range. These standards were analyzed and
the results used to assign each eluting peak to the level of chlori-

                           Table 4
           Analytical System, Fused Silica Capillary Gas
           Chromatography/Electron Capture  Detection

 DATA SYSTEM:    Shlmadzu C-R3A Chromatography Data Processor
 CHROHATOGRAPH:  Hewlett Packard 5890 Gas Chromatograph
 FEATURES:
 INLET:

 DETECTOR:

 COLUMN:

 GASES:

 TEMPERATURES:
             Split/splitless Capillary Inlet System;
             Microprocessor-controlled functions

             SpHtless

             Electron Capture

             0.25 nun I.D. X 30 n DB-5 Fused Silica (J&W Scientific)

             Carrier: Helium; 2 nl/nin   Makeup:  Argon/Methane; 30 nl/min
              Injection Port:  275 C
              Detector:       325ฐC
              Column Oven:     50-70oC 8 4 C/min
                            170-210 C 8 1 C/min
                            210-290ฐC 8 4ฐC/niin

 QUANTIFICATION:  Internal Standard: dibromooctafluorobiphenyl
              Retention Time Standards: tetrachloro-n-xylene
                                    3,4,5-tribromobiphenyl
                                    decachlorobi phenyl
                     Total PCB (as Aroclor 1242+1254+1260);
                     8.63, 2.97,  1.0 ng/ul
INITIAL CALIBRATION:


CONTINUING CALIBRATION:  Total PCB (as Aroclor 1242+1254+1260); 2.97 ng/ul
                                                                   nation determined from the GC/MS analyses. Retention time and
                                                                   pattern recognition were critical components of peak designation.
                                                                   GC-ECD peaks were further identified by relative retention time
                                                                   against the elution of three retention time markers. The use of
                                                                   relative retention  time drift enabled the chromatography data
                                                                   system to identify peaks accurately regardless of actual retention
                                                                   time.
                                                                      Peaks consisting of coeluting PCBs from different homologue
                                                                   groups were assigned the higher level of chlorination. No attempt
                                                                   was made to quantify the relative amounts of PCBs with different
                                                                   levels of chlorination of a single eluting peak for the original com-
                                                                   parability study of seawater samples.

                                                                   Results
                                                                      The results of the comparability study are presented in Table
                                                                   5. PCB concentrations are presented as level of chlorination and
                                                                   as total PCB (by summing each level of chlorination). Data for
                                                                   C1,-PCB and "C17-PCB, which often were below detection limits
                                                                   for either method, are not shown but were included in the total
                                                                   when quantifiable. The GC-ECD data were generated in this study;
                                                                   GC/MS data are taken from analyses performed earlier for the
                                                                   modeling program.
                                                                                              Table 5
                                                                   Summary of GC/MS and GC/ECD Data, Battelle Comparability Study

1
2

3
4

5

6

7

8

9

10

11

12

U

u

15

16

17

18

19

20

CC/HS
CC-ECD
GC/HS
CC-ECO
CC-ECD
GC/HS
CC-BCD
CC/HS
GC-ECD
GC/HS
CC-ECD
GC/HS
GC-ECD
GC/MS
CC-ECD
GC/HS
CC-ECD
GC/HS
CC-ECD
CC/KS
GC-ECD
CC/HS
CC-ECD
GC/HS
CC-ECD
CC/HS
GC-ECD
CC/HS
CC-ECD
GC/HS
CC-ECD
CC/HS
GC-ECD
CC/HS
GC-,ECD
CC/HS
GC-ECD
CC-ECD

•0.486 1.190
0.3996 1.5693
,403 .897
.1438 .5696
.0375
.134
.0096
.023
.0035
.009
.0138
.013
.0053
.017
.0055
ND
.0012
NO
.0015
.0022
.0010
ND
.0001
.019
.0077
.041
.0053
.006
.0046
.007
.0053
.002
.0017
HD
.0017
ND
.0005
.0006
3271
272
2696
113
0317
049
0654
066
0385
076
0804
028
0107
0056
0084
0488
0202
0193
0178
111
OB51
082
0214
046
0228
044
0334
008
0069
004
0063
007
0035
0046

.576
1.3670
.550
.4855
.7458
.255
.2673
.112
.0209
.062
.OB46
.085
.1071
.076
.10
.OB
.01
.02
.02
.12
.05
.05
.04
.10
.10
.08
.02
.05
.02
.04
.04
.02
7
3
5
7
4
6
6
7
2

9

0

5

2
6
.0205
.019
.0161
.0310
.0091
.0140

.138
.3296
.160
.1326
.2589
.093
.0847
.036
.0131
.023
.0465
.033
.0410
.033
.0373
.0480
.0219
.0173
.0221
.1103
.0492
.0318
.039
.044
.0496
.028
.0144
.013
.0413
.014
.0188
.0169
.0206
.015
.0163
.0214
.0116
.0130

MJI4"
.0352
.020
.0096
.0551
.019
.0096
.002
.0012
.003
.0053
ND
.0052
.001
.0037
.0159
.0064
ND
.0079
.0407
.0178
.0121
.0121
.005
.0057
ND
.0030
NO
.0013
ND
.0018
.0073
.0080
ND
.0058
.0051
.004
.0052

ND 3
.0042
ND
.0068
.0027
ND
.0031
ND
NO
ND
.0004
ND
.0004
ND
.0006
ND
.0001
HD
.0017
NO
.003
ND
.OOZ1
ND
.0004
ND
.0001
ND
.0001
ND
.0001
ND
.0077
ND
.0012
ND
.008
.0010
41
71
03
3511
430
774
644
286
070
146
216
197
197
203
0001
173
0588
0436
065
3306
147
1189
1163
281
2514
209
0731
116
0716
114
102
0641
0604
037
0474
065
0295
OB85
0394

.54
-33
-3

-17

-75

+47

0

*14

-66

*51

-55

-4

-10

-65

-38

-10

-o

+28

-54
-95
     CC/HS
     CC-ECD
     CC/HS
     CC-ECO
     GC/HS
     CC-ECD
     CC/HS
     CC-ECD
     GC/HS
     GC-ECD
     GC/HS
     CC-ECO
      otal_PCB (CC/ECD) - total PCB (CC/MS) xlOO
             total PCB (CC/HS)
                                                                    ND - Not detected.
  Generally good agreement is evident in the comparison of the
two sets of data. The percent deviation in total PCB concentra-
tion between the two methods ranged from -95% to +55%. There
was  some  indication that the  GC-ECD  method produced a
systematic underquantification of the less chlorinated homologues,
particularly the C12-PCB  group which was not intended to be
included in the modeling. This discrepancy was judged to be a
minor problem in the context of the modeling program considering
the otherwise close similarity of results. The enhanced sensitivity

 WATERWAYS AND  WETLANDS RECLAMATION    423

-------
of the GC-ECD method is demonstrated in several cases for which
PCB levels were below the MDL of Method 680 but were quan-
tifiable by GC-ECD.  For example, the C16 and C17 homologue
groups in sample 3 were not quantified by GC/MS but were iden-
tified by GC-ECD. No PCBs were identified in the outer Buzzards
Bay station (samples 21 and 22) by GC/MS, but GC-ECD analy-
sis of the same samples yielded quantifiable levels of most homo-
logue groups.
  Based on these results, analysis of PCB homologue groups by
GC-ECD  was determined to  be a valid analytical procedure for
water filtrate and paniculate samples containing PCB levels below
the MDL  for Method 680. Agreement between the two methods
was, in general, within a factor of two. These results are not neces-
sarily applicable to other matrices or other locations and depend,
in pan, on the combination of several factors. All field samples
contained PCB in an easily recognizable pattern allowing relatively
simple peak identification with  only one analysis. The seawater
samples and blanks contained no matrix interferences or other com-
pounds with GC-ECD response  (e.g., pesticides) that could have
complicated the PCB quantification. Although individual inter-
ference peaks were  identified  in some chromatograms, their
presence was easily determined from pattern recognition  and the
data were easily edited to remove any interference contribution.

FINAL ANALYTICAL SCHEME
  On the basis of the results of the comparability study, a proce-
dure was developed for analysis of approximately 1400 samples
determined to have expected PCB concentrations below those
necessary for quantification by Method 680. It was determined that
the method should include the following considerations to ensure
data integrity:

• A single calibration standard should be used by all laboratories
  participating in the program. The standard should contain PCB
  at all anticipated levels of chlorination, best accomplished via
  a solution of Aroclors. The identification of eluting PCB peaks
  (on a standardized capillary GC column) and the quantification
  of PCB level of chlorination must be performed by Method 680.
• A three- or five-point calibration should define  the calibration
  range. All analyses should be made within the calibration range.
  Samples containing analytes not within the calibration range
  should be concentrated or diluted, as appropriate, to meet the
  requirement.
• The relative standard deviation of instrument response during
  initial calibration must be less than 30% for the C12 through C17
  homologues.
• The difference  in response between the initial calibration and
  the continuing calibration  must be less than 30%.
• If relative retention time is  not used to identify PCB peaks, the
  absolute retention time of the internal standard must not vary
  by more than  ฑ 5  sec.
• Field samples  must be analyzed  in lots organized by matrix.
  Procedural blanks should be analyzed and potential interferences
  should be identified. Data should be edited to  remove known
  interferences.
• A second analysis of all samples on a GC column of different
  polarity must be performed. Only PCB confirmed by the second
  analysis should be quantified.

  In addition, additional comparison  of the two methods was
recommended  on samples from sediment and tissue matrices  to
ensure that the results of the seawater samples were representative.
It was recognized that both sediment and tissue samples may con-
tain higher levels of potential interferences than  water samples.
  Based on these recommendations, a method for analysis of the
remaining field samples was developed. The final strategy is shown
in Figure 3. Samples  were separated initially by matrix (e.g., bed
sediment, suspended  sediment, lobster muscle, lobster  hepato-
pancreas, etc.); each matrix type was then treated as an analytical
lot. All samples from a lot were processed according to project
protocols, including protocols for the removal of sulfur with TBA-
sulfite and protocols for removal of most polar interfering com-
pounds with alumina column chromatography. They were then sub-
jected to an initial GC screening to evaluate the magnitude of any
interferences. Samples  with  interferences  sufficient to prevent
accurate  quantification  were further processed  using alumina
column chromatography. This loop, shown in the upper right of
the flow diagram, was  repeated until  GC screening indicated
reduction of interferences to acceptable levels.







COMC-





























> tOO wt





























^i






















II TRACT
J


I
MTfmiUMCO
I"
' OCTICTt
*'
U>wf
1"
CONC
1


I
*ป*ซ

*


i-


i
OC OUAMTVCATO4 '

J
OC O*
I





P^J cuuwur



-SซJ AUOUOT
I
1 (M.UT1

—


CMC
eoซc

T



OC/ ป
i

1



                          Figure 3
         Flow Diagram for Mixed-Method Analysis of PCB
                    By Homologue Group
                          Table 6
 GC/MS and GC-ECD Data Comparison for Mixed-Method Analysis
       OC-ICT

       ce-ico
1400

1100
HOO

4*00
    r - OUปL**4 HUM'
    rtซ i  cc/m
        CC-tCO
                                           .ซ 1


                                          -li.4
424    WATERWAYS AND WETLANDS RECLAMATION

-------
  After passing the interference loop, the concentration of analyte
in the extract was compared with method detection limits for the
GC-ECD method. Extracts with low analyte concentrations were
concentrated to approximately 100  afal. Because the concentra-
tion procedure also increases interferences, it was necessary to pass
these extracts back to the screening loop as shown in the upper
left of Figure 3. Additional alumina chromatography cleanup was
conducted as necessary.  Immediately after this decision point,
volume was readjusted prior to passing the extracts on to the GC.
Quantification was performed by GC-ECD as described previously.
  For each sample matrix, two extracts were selected and analyzed
via GC/MS to derive ratiometric distributions for  encumbered
peaks. The  comparative  data for  the  three physical matrices
(seawater-dissolved  phase, seawater-particulate phase and bed
sediment) and two of the several tissue matrices are presented in
Table 6.
  The data presented in Table 6 clearly indicate the close agree-
ment between results obtained using the two methods. Data gener-
ated during actual production runs were in general more similar
than data obtained during the special comparability study. The
difference in measured total PCB between the two methods was
typically less than 20% and never exceeded 50% for the matrices
shown. Considering that achievable precision for duplicate analyses
using U.S. EPA Method 680  alone is reported to be only within
a factor of two, these results indicate exceptionally good agree-
ment between methods. Comparability between the methods does
not appear to differ between matrices with comparatively few
potential organic interferences (i.e., seawater) and those with the
potential for significant interference problems (tissues).
CONCLUSION
  Based on our results from samples collected at New Bedford
Harbor, it is apparent that U.S. EPA Method 680 is the analytical
method of choice for programs that can afford its significantly
greater cost and that either have samples that are moderately to
heavily contaminated or can collect sufficient material. For pro-
grams with limited funding or a need  for  the lowest possible
detection limits, GC-ECD analysis is more appropriate.
  For the New Bedford Program, the large  number of samples
necessary  to properly calibrate and validate the mathematical
models made funding a critical issue. Further, sample sizes had
been determined based on method detection limits for GC-ECD
analysis and were inappropriate for the lower sensitivity of Method
680. Our data indicate that a carefully planned and implemented
analytical program can use both methods and obtain results that
are sufficiently interchangeable for most purposes. In such an
analytical strategy,  GC/MS U.S. EPA Method 680 is used for
samples with sufficient PCB in extract to be detected; GC-ECD
is used for samples with PCB levels below the Method 680 detec-
tion limits. The use of U.S. EPA Method 680 as, in effect, a quality
control procedure on the GC-ECD analysis ensures  that data from
the two methods are sufficiently similar to be used interchangeably.

ACKNOWLEDGMENT
  Mr. Frank Ciavattieri, U.S. EPA Region I New Bedford Harbor
Project Manager, Dr. Alan Robbat of Tufts University and Dr.
Joseph Comeau of Aquatec, Inc., South Burlington,  Vermont, con-
tributed significantly to the development of analytical procedures
presented herein.

REFERENCES
1.  Metcalf & Eddy, Inc.,' 'Acushnet Estuary PCBs Data Management, Final
   Report," U.S. EPA Region I, Boston, MA, 1983.
2.  U.S. EPA Office of Toxic Substances, "PCBs in the United States, In-
   dustrial Use and Environmental Distribution.  Task 1, Final Report,"
   U.S. EPA-560/6-76-005, 1976.
3.  Gebhart, I.E., Hayes, T.L., Alford-Stevens, A.L. and Budde, W.L.,
   "Mass Spectrometric Determination of Polychlorinated Biphenyls as
   Isomer Groups," Anal. Chem., 1985, 2458-2463.
4.  Schwartz, T.R., Stalling, D.L. and Rice, C.L., "Are  Polychlorinated
   Biphenyl Residues Adequately Described by Aroclor Mixture Equiva-
   lents? Isomer-Specific Principal Components Analysis of Such Residues
   in Fish and Turtles," Environ. Sci. Techno!., 21, 1987, 72-76.
5.  Tetra Tech, Inc., "Bioaccumulation Monitoring Guidance: 4. Analyti-
   cal Methods for U.S. EPA Priority Pollutants  and 301(h) Pesticides in
   Tissues from Estuarine and Marine Organisms," Final Report to U.S.
   EPA, Jan., 1986.
6.  MacLeod, W.D., Jr., Brown, D.W., Friedman, A.J.,  Burrows, D.G.,
   Maynes, 0., Pearce, R.W., Wigren, C.A. and Bogar, R.G., "Standard
   Analytical Procedures of the NOAA National Analytical Facility,
   1985-1986, Extractable Toxic Organic Compounds, Second Edition,"
   NOAA Tech.  Memo. NMFS F/NWC-92, Oct., 1985.
                                                                     WATERWAYS AND WETLANDS RECLAMATION    425

-------
                 Sampling Program  for  the  New  Bedford  Harbor
                     Superfund  Site:  Modeling  the Movement of
                                PCB-Contaminated  Sediments
                                              Richard A. McGrath
                                          Michael S. Connor, Ph.D.
                                             William G. Steinhauer
                                            Battelle Ocean Sciences
                                            Duxbury, Massachusetts
 ABSTRACT
   New Bedford Harbor is one of very few estuarine Superfund
 sites. Sediments at the New Bedford Harbor "hot spot" PCB that
 frequently exceed 500 ppm and occasionally exceed  10,000 to
 100,000 ppm. Most of the routes of human exposure to PCBs are
 from contaminated sediments via fish and shellfish. Lobsters, eels
 and finfish in the region are heavily contaminated with PCBs, and
 PCB concentrations in the edible flesh frequently exceed the FDA
 action level of 2 ppm. As a part of the RI/FS, Battelle is developing
 linked hydrodynamic/sediment-transport and food-chain models
 to evaluate how two or more remedial action alternatives would
 affect the long-term fate of PCBs in the estuarine sediments com-
 pared with a no action baseline. The models consider PCBs as level-
 of-chlorination homologue groups rather than as either "total"
 PCB or Aroclors.
   The physical/chemical transport and fate of PCBs and metals
 in the New Bedford Harbor/Buzzards Bay system will be modeled
 by the time-varying three-dimensional model FLESCOT developed
 by Battelle. The model will be used to predict the distribution of
 PCBs and metals dissolved in  the water column, sorbed to three
 size classes of suspended sediments and in the bed sediments. An
 age/size-dependent food-chain model that considers bioenergetics
 of different species in the food chain and exposure to PCBs via
 the water column will use the output of the physical/chemical model
 to predict the movement of contaminants through key components
 of the food chain.
   Battelle's seasonal field sampling program to provide data for
 model calibration, validation and initial conditions was conducted
 from 1984 to 1986 in  the Acushnet River Estuary, New Bedford
 Harbor and Buzzards Bay. Approximately 4000 marine environ-
 mental samples were taken including surficial sediments, surface
 and bottom water (collected in some cases at intervals during the
 tidal cycle) and marine animals. Vertical  cores of sediment were
 collected  from stations in  New Bedford Inner Harbor. Survey
 cruises were conducted during three seasons and during a storm
 event, which was  the  most difficult  aspect of the field program
 to complete successfully.

 INTRODUCTION
   The Acushnet River estuary, which in its middle and lower
 reaches forms New Bedford Harbor, Massachusetts (Figure 1), is
 heavily contaminated with PCBs, heavy metals and possibly other
 industrial wastes. It is believed that the PCBs came from two elec-
 tronics components manufacturers  who apparently discharged
 PCB-laden wastewater directly into the harbor and the municipal
 sewer system from the 1940s  to the mid-1970s1.

426   WATERWAYS AND  WETLANDS RECLAMATION
                                                                              ro-55
tf40 -
4T35
                        Figure 1
       New Bedford Harbor Superfund Project Study Area
  Elevated concentrations of PCBs were first reported in sediments
of New Bedford Harbor in 19762. Since then, many investigations
have documented the widespread PCB contamination of sediments
and marine biota of the Acushnet River, New Bedford Harbor and
adjacent Buzzards Bay. PCB concentrations in sediments of the
upper Acushnet estuary frequently exceed 500 ppm dry weight and
occasional samples contain in excess of 10,000 to 100,000 ppm
(1 to 10%). On the local estuarine biota, eels (Anguilla rostrata)
appear to be most heavily contaminated, with body burdens occa-
sionally exceeding 500 ppm. Lobsters (Homarus americanus) also
appear to be heavily contaminated, with concentrations in muscle
tissue occasionally exceeding  50  ppm1  The Food  and  Drug
Administration (FDA) action level for the edible portion of fish

-------
and shellfish is 2 ppm and most of the area is now closed to com-
mercial and recreational  fishing.
  The New Bedford Harbor  PCB problem  was designated  a
priority in a 1980 agreement between the U.S. EPA and the Massa-
chusetts Department of Environmental Quality Engineering. In
1982, New Bedford Harbor and adjacent areas of Buzzards Bay
were designated a Superfund hazardous waste site and remedial
action planning was initiated. More recently, the U.S. Justice
Department, in conjunction with U.S. EPA, the National Oceanic
and Atmospheric Administration and the Commonwealth of Mas-
sachusetts, filed suit against the electronics components manufac-
turers to seek damages for  the loss of natural resources in the region
because of PCB contamination of resource  species and habitat.
  Under contract to Ebasco Services Incorporated, the U.S. EPA's
Superfund contractor, Battelle Ocean Sciences currently is con-
ducting a portion of the RI/FS to evaluate scenarios to remediate
some of the environmental contamination in the harbor. This pro-
gram is based on linked  hydrodynamic/sediment-transport and
food-chain models that will allow mathematical simulation of
various remedial actions.  As a first step toward addressing these
problems, a three-dimensional hydrodynamic and sediment trans-
port model is being adapted to the harbor and adjoining areas of
Buzzards Bay. After calibration and validation, this model will be
used to simulate the effects of a no-action baseline and two or more
remedial action alternatives on patterns of contaminant flux in the
area over  a 10-year period.
  The results of the hydrodynamic/sediment-transport model will
serve as input to the food-chain model. Using projected levels of
contaminants in water, sediments and suspended particulates, the
food-chain model will mathematically follow the movement of con-
taminants  through local food webs and will determine ultimate con-
tamination levels in three species eaten by humans: winter flounder
(Pseudopleuronectes americanus), lobster (Homarus americanus)
and quahog (Mercenaria mercenaria). Because consumption of con-
taminated fish and shellfish is probably the most important path-
way by which the contaminants reach humans, the model results
will be critical in evaluating  action alternatives.


DATA REQUIREMENTS FOR THE MODELING PROGRAM
  Mathematical modeling of a complex area such as the Acushnet
River/New Bedford Harbor/Buzzards Bay system requires the col-
lection, processing and interpretation of a large amount and wide
variety of data. For the Superfund modeling program, three dif-
ferent types of data were needed to calibrate and verify the models:
laboratory studies were performed to provide site-specific infor-
mation directly applicable to and required for the models; physi-
cal  oceanographic studies were conducted to determine the normal
and storm-induced patterns of water currents in Buzzards Bay and
New Bedford Harbor; and an extensive field sampling program
was undertaken to collect  data on the physical, chemical and bio-
logical characteristics of the study area for use in the model calibra-
tion and validation runs and to provide baseline data for initial
conditions.
  The field studies were initiated in the fall of 1984 and continued
into the summer of 1986.  The field sampling program proved to
be the  most costly and time-consuming aspect of the data collec-
tion effort and will be discussed in detail following a brief over-
view of the two  other program components.

Laboratory Experiments
  Two types of laboratory experiments were conducted to provide
site-specific values for critical model parameters.  A partitioning
experiment  was  performed  to provide adsorption/desorption
kinetics and partition coefficients (Kds) for PCBs and metals in
bed sediments and suspended sediments. The  Kds and other
physical parameters determined empirically will be more accurate
(relative to local conditions) than literature values and will better
reflect the unique physical/chemical properties and contamination
of the harbor sediments.
  Laboratory studies also were performed to determine bioaccumu-
lation factors, uptake efficiency from food and rates of release
of different PCB components and metals by the food-chain species
of concern, namely lobster, flounder  and clams. In both experi-
ments, PCBs were analyzed by level-of, chlorination homologue
group rather than as total PCB or as Aroclors.

Collection of Hydrographic  Data
  The major feature of the oceanographic data collection was the
deployment of three arrays of current meters at the Buzzards Bay
boundary (Figure 2).  At each station, the array consisted of two
current meters at 0.8X and 0.2X water depth, nominally defined
as bottom and surface, respectively, for the purposes of the model.
Two  separate  deployments, of approximately  6 months in
1984-1985 and 9 months in 1985-1986, were necessary to encom-
pass the  three scheduled field samplings and one  storm event
sampling. These deployments established boundary tidal forcing
conditions in summer and winter for the  coarse model grid used
over most of Buzzards Bay. The coarse grid will be used to drive
a finer grid encompassing most of New Bedford Harbor and the
Acushnet River Estuary.
                          Figure 2
        Buzzards Bay and Bay Boundary Station Locations
  Additional shorter-term current meter and drogue studies were
performed in outer New Bedford Harbor to establish patterns of
current flow in this region. Three separate drifter studies were con-
ducted in November and December of 1984 and April of 1985.
During each study, surface and subsurface drifters were released
in the outer harbor and tracked via radiotelemetry for 3 to 5 days.

 WATERWAYS AND WETLANDS RECLAMATION    427

-------
In addition to providing data for later comparison with model-
derived streakplots, the comparison of surface and subsurface
tracks yielded information on wind-induced shear in the top few
meters of the water column. A fourth  drifter study, primarily in
the inner harbor, was undertaken in the summer of 1986 to provide
empirical data on wind and tidal forcing of currents in the area
of prime importance for assessing the effects of remedial actions.
The inner harbor study also provided data on water dispersion and
residence time in the harbor.

FIELD SAMPLING PROGRAM
  The field sampling program comprised 25 stations for chemis-
try sample collection, eight of which ("primary stations") were
sampled more intensively than the other 17 and four "areas" for
collection of biota. The stations ranged from the uppermost  part
of the Acushnet River estuary to the easternmost and westernmost
parts of Buzzards Bay; station  density was highest in the inner
harbor, the area of most  elevated PCB contamination and com-
plex tidally driven currents. Station locations were chosen to include
the hydrographic boundaries of the region, potential sources of
contaminants and areas important for characterizing the circula-
tion in the Harbor (Figures 2 and 3). To allow predictions to be
relevant to the regulatory boundaries  in  New Bedford, the four
biota areas were selected to correspond with the different fishery
closure areas established by the Massachusetts Division of Marine
Fisheries.
               55'
                        70*50
                                               700ซ0'
4Iซ40' -
4Iฐ30' -
   25' -
                          Figure 3
     New Bedford Harbor and Acushnet River Station Locations
  Collections were made on three scheduled seasonal surveys in
 1984 and 1985 at 25 stations and during an additional storm event
 sampling that was conducted in 1986 at the eight primary stations
 (stations 3,9,10,12,16,18,21 and 24; see Figures 2 and 3). Samples
 collected on the scheduled surveys included surface and bottom
 seawater,  bed sediment and several species of biota. At the eight
 primary stations, seawater collections were made at four times
 during  a tidal cycle; the remaining 17 secondary stations were

428     WATERWAYS AND WETLANDS RECLAMATION
sampled only once during a survey. Additional measurements were
made of current velocity, temperature and salinity.
  Most samples collected during this phase of the program were
separated into two or more fractions. Bed sediments were sepa-
rated for analysis into sand, silt, clay and pore water fractions
(Figure 4). Vater samples were separated into paniculate  and
dissolved fractions (Figure 5). For flounder and lobster, edible por-
tions (muscle of flounder, muscle and hepatopancreas of lobster)
of the biota samples were removed  for separate analysis. These
separations were necessary because the contaminants in the dif-
ferent fractions behave differently, both physically and biologically
and the models are designed to consider and predict these different
behaviors.
                                                                                            Figure 4
                                                                              Immediate Processing of Sediment Samples
                          Figure 5
            Immediate Processing of Seawater Samples
Seawater Sampling Methods
  Water samples were collected using a non-contaminating peristal-
tic pump equipped with Tenon tubing. The tubing was attached
to a measured nylon line and lowered to the required  sampling

-------
depth. Flow rate through the pump was approximately 1 1/min.
To minimize cross-contamination, the pump inlet was placed at
the desired depth and the system was flushed for 30 min between
samplings.
  At each station, a suite of samples was collected from the surface
and bottom (0.2X and 0.8X  depth,  respectively). Each  suite of
samples comprised three 11 samples for metals' paniculate organic
carbon (POC) and total suspended solids (TSS), respectively and
a 20-1 sample for PCBs. The  FCB sample vas filtered through a
ISO-mm Millipore filter at the time of collection, yielding  a parti-
culate and dissolved  fraction. The metals  sample was similarly
filtered upon return to the laboratory.

Sediment Sampling Methods
   Sediment  sampling was  performed using an 8.4 cm diameter
hydraulically-damped, gravity corer made available by the U.S.
Geological Survey, Woods Hole, Massachusetts (Figure  6). The
hydraulic damping action slows the penetration of the corer into
the sediment and minimizes disruption of the high-organic floc-
culent layer, believed to be high in PCB contamination, at the
sediment-water interface.,This effect is further enhanced because
the corer does not require a nosecore and uses core tubing with
a narrow (0.24 cm) wall. These characteristics allow the corer to
penetrate with little disruption of sediment integrity, thus  making
it ideal for the collection of both undisturbed surface material and
core samples to 50 cm. By using a clean core barrel with each lower-
ing, cross-contamination caused by the presence of contaminated
materials from hot spots was  minimized. In most cases, only the
top 5 cm of sediment were analyzed; at three stations, samples were
collected  from two additional S-cm strata to examine PCB distri-
bution in the layers most affected by  bioturbation.
 1
Biota Sampling Methods
  Six taxa of marine animals were collected for chemical analy-
sis. These included the three species of primary interest as a path-
way  to  seafood consumers  (American  lobster,  Homarus
americanus; winter  flounder, Pseudopleuronectes  americanus;
quahog, Mercenaria mercenaries) and three additional taxa of
importance in local  food chains (spider crab, Libinia sp.; blue
mussel, Mytilus edulis; and mixed species of polychaete worms).
Specimens in various size classes (Table 1) were collected in each
of the four areas on each  of the three scheduled surveys.
                            Table 1
              Biota Samples and Collection Methods
                    For the Food-Chain Model
      Species
                                               Collection Method
                           <75 nn;75-105 M;>105mm   Lobster Traps
                            (carapace length)      Otter Travl
Lobster
(Hoaarus americanus)

Winter Flounder            <17B ซซ;i78-254ซซ;>25i*ป   Otter Trawl
(Pseudopleuronectes americanus)
    Rock Crab (Cancer irroratus) or <20 am; 20-50
    Spider Crab (Libinia enarglnata)
    Hard-Shell Clam
    (Mercenarta aercenarla)
    Mussels (Hytilus edulis)
    Incaunal Polychaete
    (Hephthys, Clycera)
                       <30mmi 30-60 ซซ|>60 mm


                       No Size Classification

                       Ho Size Classification
Otter Travl


Rocking Chair Dredge


Band Collection

Crab Sampler
                              Figure 6
               Hydraulically Damped Corer Provided By
             U.S. Geological Survey, Woods Hole, Mass.
  Collection gear was varied and included commercial-type trawl nets and
dredges for all species except the mussels, which were collected by hand
from intertidal sites and the polychaetes, which were collected with Smith,
Mclntyre or ponar grab samplers. All biota were washed with site seawater
after collection, wrapped in acid-washed aluminum foil, bagged in poly-
ethylene and frozen in the field.

ANALYTICAL CHEMISTRY PROGRAM

  For all samples, PCBs were analyzed as homologue or pseudo-component
groups according to their level of chlorination. All 10 possible such groups
were quantified, but the model will consider  only the C13 through C16
homologues, which were the major constituents of the PCB blends used
at New Bedford and are the primary components of the PCBs remaining
in the harbor sediments. Standard atomic absorption (AA) techniques were
used to analyze metals samples for copper, cadmium and lead. Analysis
of other samples (e.g., sediment grain size, TSS, POC) followed routine
accepted procedures.

QUALITY ASSURANCE PROGRAM

  To ensure that the samples reflected environmental levels and not artifacts
of collection, rigid protocols for sample collection and shipboard processing
were developed for important survey operations. Sampling protocols speci-
fied shipboard conduct and precautions, methods of navigation, positioning,
sampling and sample processing. In addition, because the program require-
ments specified adherence to chain-of-custody procedures, similar specifi-
cations were  developed for the recording, processing,  reviewing and
reporting of data. Period performance audits were conducted by Battelle's
Quality Assurance Unit to evaluate adherence to the protocols.
  Because the sample analysis was the responsibility of a separate subcon-
tractor and many  samples were not expected to be analyzed soon after
collection, it was most  important to  develop protocols specifying how
samples should be processed for indefinite storage. For example, sediment
samples requiring pore water analysis were processed in a glove box under
nitrogen to preserve the sediment in its anoxic state. Ideally this work would
have been conducted on board ship in controlled laboratory conditions;
however, a vessel large enough to contain a clean room and laboratory
facilities would not have been able to navigate properly in the study area.
To avoid the substantial additional costs of processing the samples in a
shore-based, portable laboratory, a "runner"  met the ship at the dock to
accept the samples according to a strict chain-of-custody procedure, drove
                                                                         WATERWAYS AND WETLANDS RECLAMATION    429

-------
them 30 miles to our laboratory facilities and transferred custody to the
laboratory. The samples were then processed to a point at which they could
be stored before analysis.
  Quality control samples also were taken on each survey, accounting for
approximately 5% of the total samples analyzed. These samples were of
four types:

• Shipboard contaminants, e.g., atmospheric deposition, bilge water, fuel
  oil and hydraulic fluids
• Duplicate samples from an uncontaminated sampling area (Duxbury Bay,
  MA) to check for contamination or bias introduced by the sampling
  equipment or techniques
• Split samples for re-analysis to determine analytical and sampling varia-
  bility
• Laboratory-spiked samples to verify the accuracy of analytical procedures

CONSIDERATIONS FOR  THE DESIGN OF
FIELD SAMPLING TO SUPPORT MODELING
  Field sampling can serve many purposes, including baseline assessment,
compliance monitoring and risk assessment. Because the purpose of this
sampling program was to support the development of a model to predict
the effects of different remedial actions on the distribution and long-term
concentration of PCBs and metals in New Bedford Harbor waters, sedi-
ments and  fisheries, special problems in sampling design arose.
  To obtain values that  were of maximum usefulness for calibration and
validation of the models, water chemistry, sediment chemistry and physi-
cal oceanography data had to be collected as synoptically as possible. The
optimal number of stations  for modeling purposes initially was estimated
to be over 100, a number thai would have precluded temporal synopticity
and  would have far exceeded  the available funding. By balancing the
modelers' requests, the capabilities of the available ships and the budge-
tary  concerns, agreement  was reached on the  smaller  set  of sampling
stations.
   To further increase the cost-effectiveness of the field program,
a rigorous logistical plan was implemented  to optimize time spent
in the field. As was  noted above, some stations required tidal
sampling, but others could be sampled at any tidal stage. The cruise
plan for each survey considered the time of occupation of tidal
stations first and then, factoring in the vessel cruising time, filled
in the remaining stations to eliminate  nonproductive time in the
field. The final plan took into account  not only those  factors, but
also the requirement  to sample at the  less  contaminated stations
first; times of scheduled bridge openings to allow optimal transit
between the harbor and estuary; scheduled vessel maintenance; and
the need to transfer samples to the laboratory, which was operating
on its own 24-hour schedule, for immediate processing and storage.
The tight scheduling required  some revision in the field due to
unanticipated failure of the vessel and sampling equipment but
generally  was workable  and cost-effective.
   The most difficult modeling requirement proved to be the neces-
sity for sampling during a storm event. The resuspension of con-
taminated sediments during storms is thought  to be the  major
transport mechanism  for contaminants  from this site. Planning for
a storm cruise required, first, consensus among the modelers, physi-
cal oceanographers and sampling team on the definition of a storm.
The modeling team not  only requested data from a large storm,
but also wanted data as near the height of the storm as possible.
These requirements were mutually exclusive because of the small
size of the sampling vessel (13m), a decision that was in turn dic-
tated by the need to navigate in shallow and confined areas of the
harbor. A storm was eventually defined as a sliding scale of wind
and  duration ranging from 40 +  knot winds for at least 4  hr to
30 knot winds  for 12 hr.
  In practice, these criteria were  nearly impossible to apply. It
would have been prohibitively expensive to keep the field sampling
team on standby and, even though our laboratory is approximately
a 1-hr  drive from New Bedford, a full day advance notice was
required to  mobilize the team and equipment. The only period
during which storms of the required intensity are probable in New
England is winter, when prediction is more difficult and the harbor
is often frozen. These difficulties, combined with two successive
winters (1984-85 and 1985-86) of unusually mild  weather, pre-
vented the collection of the storm event samples for  well over 1  year.
  As the program progressed, it became apparent that a more sub-
jective definition of a storm would be necessary'  It was agreed that
team members would make an ad hoc decision during subsequent
periods of strong winds. Eventually, a period of high winds (20
to 30 knots) but otherwise clear weather  occurred  from Feb. 25
to Feb. 26, 1986. Although these conditions did not meet the  strict
definition of a storm, the project team decided that there was suffi-
cient sediment transport occurring in the harbor and bay to pro-
vide useful  data for the  model  and the storm sampling was
completed.

CONCLUSION
  Field sampling programs in support of fate and transport models
require careful  balancing of model needs with logistical and cost
considerations. Current budgetary limitations make it increasing-
ly unlikely that modelers w ill be provided with the full suite of data
they would prefer in order to calibrate and validate their models.
Thorough examination  of data needs when planning a program
can help produce a field  sampling program that provides sufficient
information at  acceptable cost.  To  ensure the  success of this
process,  modelers,  field  personnel and  managers must work
together  and be sensitive to the unique constraints facing  each
group.
  In the case of the New Bedford Harbor modeling program, field
sampling costs were controlled through effective interaction among
program participants in the design of the sampling and careful
logistical planning of time in the field. The storm event sampling,
however, did not allow this type of planning and ultimately became
one of the most difficult requirements of the program to complete.
Similar investigations should carefully evaluate the  need for such
data to ensure  that limited funds  are allocated effectively.

REFERENCES
1.  Metcalf & Eddy, Inc.,  "Acushnet Estuary  PCBs Data Management,
   Final Report," U.S. EPA Region I, Boston, MA,  1983.
2.  U.S. EPA Office of Toxic Substances, "PCBs in the United States, In-
   dustrial Use and Environmental Distribution. Task  1,  Final Report,"
   EPA-560/6-76 005, U.S.  EPA, Washington, DC, 1976.
430    WATERWAYS AND WETLANDS RECLAMATION

-------
             A  Wetland Assessment  Procedure for  Superfund  Sites

                                                     Anne Sergeant
                                         Alliance  Technologies Corporation
                                                Bedford, Massachusetts
ABSTRACT
  Wetland assessments must be performed at Superfund sites where
remedial activities are expected to alter the ecosystems. The purpose
of these studies is to determine the functional values (e.g., wild-
life habitat, groundwater recharge and discharge, flood storage and
critical habitat for endangered species) of impacted wetlands.
  Superfund sites present unique challenges in wetland assessment
because in many cases their natural ecosystems have been dra-
matically altered by disposal of hazardous waste. Adjacent wet-
lands also may  have  played  an important role  in  pollution
attenuation, protecting nearby ecosystems from the effects of con-
tamination. In addition, the impacts of proposed remedial activities
such as landfill capping or removal of contaminated materials must
be determined, and it may be necessary to implement mitigative
measures such as replacement of wetlands that have been disturbed
by disposal or remedial actions.
  This wetland assessment procedure examines ecosystem, wild-
life habitat, hydrologic, water quality and socioeconomic functions.
It also determines the extent and current impacts of contamina-
tion, evaluates the potential impacts of remedial alternatives and,
where necessary,  develops  appropriate mitigative measures. This
information is critical for  preparation of Endangerment  Assess-
ments and Feasibility Studies conducted under Superfund for any
site that contains a wetland.
  For a 70-acre Superfund site in Massachusetts that generates an
estimated 33 million gallons of leachate per  year, this procedure
was used to determine the functional values of, and impacts to,
14 surrounding  wetlands  as well as the pollution  attenuation
capacity of a large cattail marsh. These data were used to develop
mitigative measures  that minimized the adverse impacts of the
proposed remedial action (capping with a synthetic membrane) on
adjacent wetlands while reducing the volume of leachate generated
by 99%.

INTRODUCTION
  Because wetlands traditionally were regarded as useless land, they
became the preferred location for garbage dumps and, eventually,
sanitary landfills. Until recently, there has been little concern for
the eventual fate of the materials disposed there. At one time certain
wetlands, if they were formed by a naturally impervious layer such
as a clay deposit, were considered relatively  good disposal areas
because they would restrict the migration of wastes. However, if
wetlands  with complex regional or local hydrologic connections
are used  for waste disposal, contamination  in one wetland can
spread to other wetland areas and aquifers.
  The U.S. EPA is charged with protection of public health and
welfare and the environment but historically has directed much
of its resources toward public health concerns. Recently, however,
there has been a greater focus on protection of the environment
on ecological grounds. Consequently, investigations at Superfund
                           Table 1
                  Steps in a Wetland Assessment


       PREPARATION
            Working Map  of Wetlands
            Background Materials

       FIELD INVESTIGATION
            Verify Wetland Boundaries
            Collect Field Data
                 Ecosystems
                 Wildlife Habitat
                 Hydrology
                 Water Quality
                 Socioeconomic
            Record Observed Impacts

       DATA INTERPRETATION
            Evaluate/Rank Wetland Functional Values
                 Ecosystems
                 Wildlife Habitat
                 Hydrology
                 Water Quality
                 Socioeconomic

       EVALUATION OF CURRENT IMPACTS
            Chemical Impacts
            Physical Impacts

       EVALUATION OF REMEDIAL ALTERNATIVES
            Effect on Wetland Functional Values
            Develop Mitigative Measures
sites may require a wetland assessment to determine the impact
of contamination on the environment.
  A wetland assessment is generally one step in the RI/FS process.
Although  it includes information germane  to  RIs,  FSs and
Endangerment Assessments (EAs) (Fig. 1), it usually is prepared
as a separate document. A wetland assessment may also be per-
formed to evaluate the impacts of remedial measures already taken
at a site. In any case, its purpose is to appraise wetlands at a site
in terms of their functional values, to determine the environmental
impacts of contamination and remediation and to develop recom-
mendations to avoid or minimize these impacts.

PREPARATION
  The procedure begins with a review of existing data for the site
to determine the location of the wetlands. Maps from the National
Wetlands Inventory (NWI), the USGS, Soil Conservation Service
(SCS) and local government are  the most useful sources. Older
editions of maps can  be helpful if they show features that have
since changed, such as an area that has been mined for gravel or

  WATERWAYS AND WETLANDS RECLAMATION    431

-------


WETLANDS PRESENT REMEDIAL
INVESTIGATION
1


CHARACTERIZATION ENDANGERMENT
t CURRENT IMPACT ASSESSMENT


EVALUATE ALTERNAT I VES : FEAS 1 B 1 L 1 TY
ซ COMPLIANCE WITH REGULATIONS STUDY
* 1MPACTS/MIT1GATIVE MEASURES



SELECT/REFINE RECORD OF
REMEDY DECISION

<

SITE RESTORED
                           Figure 1
               Wetland Assessments and Superfund.
 a wetland that is now filled in. Aerial photos, if available, are
 especially valuable because they can provide historical data for com-
 parison, particularly in cases where filling or other physical alter-
 ations have occurred. Data gathered as pan of an RI can be helpful,
 too.
   If the site has not been surveyed, USGS maps can provide a base
 map onto which wetland boundaries from the corresponding NWI
 map can be overlain. Wetland soil types shown on local SCS maps
 can be used to refine the study area and to determine boundaries
 if an NWI map is not available. In some areas, local zoning maps
 include wetland boundaries, although these maps are of variable
 quality and accuracy. Finally, aerial photos can help pinpoint areas
 of special concern. Once evaluated and summarized, these data
 are used to prepare a map of the area for use in the field and to
 select potential investigation or sampling points.
   Geologic maps and hydrologic data are also reviewed to select
 additional field observation points. For example, geologic features
 such as fracture zones and regional groundwater flow data (e.g.,
 direction and rate) can be used to predict where water-soluble con-
 taminants might be transported and thus focus the investigation
 toward  such areas.
   Other preliminary work includes tabulation of plant and animal
 species expected to occur in the vicinity of the site based on data
 from State and Federal agencies, organizations such as universi-
 ties and conservation groups, and desk references. Data regarding
 endangered species, including designated critical  habitat areas, can
 be obtained  from  the U.S.  F&WS  and state Natural Heritage
 Programs. These lists are used in the field as checklists that may
 be easily summarized and evaluated later.

 FIELD  INVESTIGATION
   In the field, wetland boundaries that have been plotted on the
 working wetland map are verified and modified  as necessary. The
 U.S. EPA defines wetlands by  three criteria:

 •  A predominance of obligate or facultative wetland plant species
   (as classified by the U.S. F&WS1)
 •  Hydric soils (those formed under wet conditions, including his-
   tosols [organic soils] and soils in  aquic suborders)
 •  Hydrologic regime

432    WATERWAYS.AND WETLANDS RECLAMATION
   All three of these conditions should be met for the U.S. EPA
 to classify an area as a wetland.
   In many cases, the presence of wetland plant species provides
 sufficient information to determine the wetland boundaries. Docu-
 ments such as Cowardin, et al.2 are also helpful in determining
 wetland types. However, if there is any doubt as to whether a par-
 ticular location is within a wetland, at least one soil pit is dug to
 verify the boundary by the presence or absence of hydric soil such
 as muck, peat or marl. The soil also is examined for evidence of
 wetland hydrologic regime such as mottling (patches of red oxidized
 and grey reduced iron) or gleying (grey color from reduced iron).
 If needed, site-specific hydrologic data are also used to verify wet-
 land boundaries.
  If the Wetland Assessment is being conducted as part of an RI,
 the boundaries may be flagged and incorporated into the site map
 by the survey team. If the mapping has been completed, the loca-
 tions of wetland boundaries are noted in the field and later added
 to the site map.
  The vegetation at each observation point is quantitatively evalu-
 ated in terms of% cover in each stratum (tree, shrub, herb and
 emergent) and overall Voage of wetland species; productivity, spe-
 cies composition and diversity and suitability for wildlife habitat
 (food, cover and  breeding area) are ranked on  a qualitative
 high/medium/low  or  good/fair/poor scale. Observed plant and
 animal species are  recorded.
  Depending upon the site, it may be desirable to survey open water
 areas for benthic macro-invertebrate population density and diver-
 sity. In addition, water temperature, dissolved oxygen and pH may
 be measured. If it becomes apparent that other  areas are worth
 investigating (for example, a peripheral area or one too small to
 be shown on available maps),  these observation points are added
 in the field.
  Each observation point is qualitatively assessed in terms of hydro-
 logic functional values including: groundwater recharge and dis-
 charge;  water purification and   filtration;  flood  storage and
 retardation and shoreline protection. Any sign of recreational or
 socioeconomic use, such as hunting or fishing, also is noted.
  Finally, the area is examined for  evidence of environmental
 impacts. These impacts can include: leachate seeps; areas of stressed
 or dead vegetation; drainage alterations; severe erosion or sedimen-
tation; and the presence of opportunistic plant or animal species.
 Monocultures of certain plants such as the reed Phragmites may
 indicate ecosystem stress. If possible, the origin of disturbance
 (physical or chemical) is  determined.


 INTERPRETATION  OF FIELD  DATA
  The data collected in the field are first summarized and examined
 for trends. The locations of wetland boundaries are transferred
 onto the site map if this has not been done by the survey team.
 Summary tables of plant and animal species noted in the field are
 prepared. Observation points  are then evaluated in terms of wet-
 land functional values3 which are grouped into  five categories:

 • Ecosystem  Functions—Primary production,  decomposition,
  nutrient export and nutrient transfer
 • Wildlife Habitat Functions—Density and number of vegetative
  strata, floral diversity, "edge" (the transition  between two
  plant  communities) and food production
 • Hydrologic Functions—Groundwater recharge and discharge,
  flood storage and retardation and shoreline protection
• Water Quality Functions—Sediment  trapping,  oxygen pro-
  duction, nutrient storage and removal and contaminant storage
  and removal
 • Socioeconomic Functions—Aesthetic, recreational, educational,
  scientific, historic and economic values

  Each observation point or wetland area is ranked in terms of
relative  functional  value  using  a  high/medium/low  or
good/fair/poor hierarchy. It is  inappropriate  to use numeric

-------
rankings because this process invites comparison of unrelated
values, such as flood storage and wildlife habitat.

EVALUATION OF CURRENT IMPACT
  Evaluation of current impact is a feature unique to wetland
assessments at Superfund sites. However, it is often difficult to
determine whether observed impacts are due to physical changes
(such as filling, draining and regrading) or chemical contamina-
tion. Generally, physical impacts are more easily observed. Some,
such as sedimentation, are quite obvious while others, like drainage
alterations, are more subtle because  their effects may manifest
themselves very slowly. Many times the vegetation at a site provides
little indication of change. For example, Atlantic white cedar
(Chamaecyparis thyoides) is extremely sensitive to alterations in
hydrologic regime and even small changes in drainage may be fatal,
although other species in the same area are unaffected.
  Chemical contamination is more difficult to discern, because its
effects are often similar to those induced by other plant stresses.
For example, stressed vegetation near a highway that runs past a
site may be  suffering  from  exposure to ozone, road salt or
hazardous substances. In addition, some plants show little sensi-
tivity to contamination: cattails (Typhaspp.), a common wetland
species, are quite tolerant of pollutants,  including heavy metals,4'
5> 6 and can even grow in leachate seeps.

EVALUATION OF REMEDIAL ALTERNATIVES
  Wetland assessment at a Superfund site usually calls for evalu-
ation of remedial alternatives.  Such alternatives may range from
capping the contaminated  area, with  only minimal impact to
adjacent wetlands,  to  removal  of all contaminated  material,
including the wetlands. Another possible  alternative is a pump-
and-treat facility that removes  contaminated groundwater, treats
it and returns it to a groundwater or surface water system.
  The  recommended alternative should be consistent with U.S.
EPA policy as promulgated in its 404(b)(l) guidelines and Execu-
tive Order 11990. Normally, avoidance of impacts  is preferable
to mitigation. But in the case of Superfund sites, damage usually
has already occurred, and  virtually  any  mitigation will  be an
improvement over current conditions. However, it is still possible
to evaluate alternatives in terms of their potential alterations of
wetland functional values.
  Most remedial alternatives will have both positive and negative
effects on wetland functions. For example, a pump-and-treat
facility may improve water  quality and at the same time drain a
wetland, or construction of a sedimentation basin may eliminate
                           Figure 2
          Superfund Site, Landfill and Associated Wetlands
nutrient and sediment loading to receiving waters while reducing
wildlife habitat. Clearly, some losses will likely be associated with
cleanup, but the recommended remedial alternative should avoid
impacts to the extent practicable and provide compensation for
lost wetlands on a one-for-one basis.

CASE STUDY
  Fourteen wetlands were located within the study area for a
70-acre Superfund site in Massachusetts (Fig. 2). This site began
as a burning dump in the 1950s and ultimately became a municipal
landfill that accepted chemical  waste. Off-site migration of vola-
tile organic compounds and metals contaminated and forced the
closure of bedrock wells at a nearby condominium complex. In
addition, contaminated groundwaters and surface waters, landfill
runoff and leachate seeps contaminated two streams and several
nearby wetlands.
  As of 1985, the landfill contained about 4 million yd^ of material
and was estimated to generate over 33 million gal of leachate/year.
Wetlands  1, 3, 5, 6, Al  and A2 (Fig. 2) exhibited obvious signs
of contamination (e.g., dark leachate or stains, oil films amd chemi-
cal odors), although plant stress was observed only in Wetland 6.
Although dead vegetation was observed in Wetlands 1,  2 and G,
there was no evidence of contamination; the plants probably died
as a result of hydrologic changes. Sedimentation, a strictly physical
impact, had occurred in Wetlands 2, 5 and G. Wetlands C, D and
F showed  no  sign of contamination or other impacts associated
with the landfill, and it appeared that contamination reached Wet-
lands  4 and E only occasionally, if at all.
  Although all 14 wetlands were evaluated in the assessment con-
ducted for this site, only Wetland 4, a 2-acre wet meadow just south
of the landfill, will be examined in detail for this case study.

Field  Investigation
  Wetland 4 contains no open water and  drains via a culvert under
the road, which surfaces in a depression in a cow  pasture and dis-
charges into Dunstable Brook. It receives landfill runoff, but only
after it has passed through and been filtered by  Wetlands 2 and
3. No signs of contamination or vegetative stress were observed.
Thus, it does not appear to be  directly impacted by the  landfill.
The eastern portion of the meadow, which receives surface water
flow from Wetland 2, is vegetated with sensitive fern (Onoclea
sensibilis), meadowsweet (Spirea latifolia),  steeplebush (S. tomen-
tosa), asters (Aster spp.), Sphagnum moss and several species of
goldenrod (Solidago spp.), grass and sedge.
  A soil pit revealed Scarboro muck with  a 4-in.  layer of decom-
posed black organic matter underlain by about 12 in.  of dark
greyish-brown, mucky sandy loam. There were  lenses of coarse
sand and gravel, indicating past deposition  during flooding. Below
this was an abrupt transition into  mottled (containing  areas of
orange oxidized iron and grey reduced iron) yellowish silt. The soil
was quite  saturated considering the time of year (October): in 2
to 3 min, 4 or 5 in. of water seeped into  the hole and the sides
rapidly collapsed.
  The wetland's long axis is inhabited by grasses and sedges. The
western portion of the meadow, into which Wetland 3 flows, con-
sists of hummocks and is very densely vegetated with purple looses-
trife  (Lythrum  salicaria), climbing nightshade (Solarium
dulcamara), and several goldenrod and sedge species. This part
of the wetland is considerably wetter underfoot than other sections.

Interpretation of Field Data
Ecosystem Functions
  This wetland is ranked high for primary production because it
provides abundant food for wildlife,  particularly birds. Organic
matter is broken down and made available at a moderate rate, so
decomposition is ranked medium. However, because flow through
the meadow is slow, most nutrients do not leave the area via surface
water; thus it is ranked low for nutrient export. On the other hand,
nutrients incorporated into plant materials are available to herbi-
                                                                   WATERWAYS AND WETLANDS RECLAMATION    433

-------
vores and cycle through terrestrial and aquatic food webs. Because
much of the food produced here is consumed by mobile species
(birds), nutrient transfer is  high. Overall, Wetland 4 is ranked
medium for ecosystem functions,  primarily because it does not
export nutrients or  support  aquatic organisms.

Wildlife Habitat Functions
  This area is a wet  meadow, unlike nearby wetlands. While only
the shrub and herb/emergent strata are occupied, they are quite
dense. A wide variety of plants provides  food, cover and repro-
ductive resources for a similarly wide variety of animals. "Edge,"
the transition between vegetative communities (in this case, from
forest to wet meadow) is an especially valuable zone to wildlife.
This zone is quite diverse and provides abundant  food and cover.
  Soil characteristics (i.e., mottling) show that the meadow exhibits
another transitional phase, a seasonal fluctuation between saturated
and unsaturated conditions, so it provides additional habitat for
species  that require  wet conditions for only part of the year. The
plant community produces a variety  of seeds, fruits and browse
and can support a reasonably-sized consumer population, which
will in turn support predators. Wetland 4 receives high rankings
for density and number of vegetative strata, floral diversity, "edge"
and food production,  and overall  wildlife habitat functions.

Hydrologic Functions
  Wetland 4 is primarily a discharge area where groundwater moves
upward and/or laterally into surface water. It receives precipitation,
surface water and groundwater flow from  the landfill and vicinity.
Because it is relatively wide and has a constriction at its outlet,
it has moderate to high capacity for flood storage and retardation.
There is no open water associated with this wetland, thus  it has
no  value for shoreline  protection.  Its overall ranking for overall
hydrologic functions is low-to-moderate.

Water Quality Functions
  There are no clear channels here and the vegetation is dense,
thus the velocity of surface water is reduced when it enters Wet-
land 4. Because water's ability to carry sediment decreases with
velocity, this area acts as an effective sediment trap. The meadow
receives full sun for  most of the day, providing ample opportunity
for photosynthesis.  Thus, its plant community is capable of high
oxygen production.
  The vegetation and  microorganisms, through  their metabolic
processes, also effectively  remove and  store nutrients such as nitro-
gen and phosphorus, which reduces the possibility  of algal blooms,
excess macrophyton  production, and associated eutrophic processes
in receiving waters.  Removal  of particulates  and nutrients also
reduces  biochemical oxygen demand  (BOD) and  downstream
pollution.  However, much  of the water Wetland 4 receives is
probably "pre-filtered" by  Wetlands 2 and 3, so its  ranking is
reduced to medium. Although it has a high capacity for  water
quality  improvement, its relatively small size reduces its overall
ranking to medium-to-high.

Socioeconomic Functions
  Wetland 4 is somewhat secluded and surrounded by forest. It
contains a number of flowering species and attracts a variety of
animals, so it is ranked high for aesthetic value. Because of its
habitat value (particularly "edge") and proximity to several other
habitat types (forest, pasture), this wetland makes an excellent area
for birding and other wildlife observation  but is not of interest for
most other recreational activities and so receives a medium ranking
for this function.
  Although the meadow is  not a  unique ecosystem,  it is  easily
accessible and may  be of some  value for ecological  study. No
historic or economic values are known for the area. Overall, it is
ranked medium for socioeconomic values.

Evaluation of Current Impact
  The meadow does not appear to have been impacted by disposal
activities or contamination associated with the landfill. As discussed

434    WATERWAYS AND WETLANDS RECLAMATION
above, runoff from the landfill must pass through Wetland 2 or
Wetland 3 before it reaches Wetland 4. Apparently, these wetlands
have adequate filtration, flood storage and sediment-trapping
capacity to protect the meadow, as no signs of contamination or
stress were observed. These two wetlands may also serve to filter
groundwater before it enters Wetland 4.

Evaluation of Remedial Alternatives
  The U.S. EPA selected a source control remedial alternative con-
sisting of a full synthetic membrane cap with leachate collection,
gas collection and venting, surface water collection and diversion,
and regrading as needed to stabilize slopes. The selected remedy
would control runoff and erosion and  was predicted to reduce
leachate production by 99%.
  As originally planned,  installation of the perimeter drain and
associated regrading would  have  obliterated  Wetlands 1 and G
through filling and drainage. The plans also  called for portions
of Wetlands 2, 3 and  B to be dredged for drainage purposes. In
addition, most of Wetlands 3 and 4 would have been altered either
by channelization or by installation of a sedimentation basin. Chan-
nelization would also  have reduced Wetland 4's ecosystem, wild-
life habitat, pollution attenuation and hydrologic  functions.
  Based on evaluations of wetland functional values and the recom-
mended remedial alternative, measures were developed to avoid
impacts, minimize unavoidable impacts or compensate for lost wet-
lands. It was recommended that relocation of the sedimentation
basin be considered in order to  avoid direct impacts to Wetlands
3 and 4.
  Because  relocation  of  the perimeter drain was not feasible,
expansion of Wetland 2 and creation of additional wetlands was
proposed. The sedimentation basin was considered a likely candi-
date for replication once the landfill cap had been stabilized with
vegetation. (Replication, the creation of a wetland similar to one
that has been destroyed, would have consisted of dredging  the
sedimentation basin to a depth sufficient to support wetland vege-
tation and planting emergent species such as cattails.) Also recom-
mended was the use of existing wetlands for storm water retention.
Erosion and  sedimentation controls were suggested to augment
those already planned for the construction process.
  Finally, the assessment proposed a monitoring program designed
to observe changes in wetland ecosystems and leachate generation.
This monitoring is critical not only to evaluate the changes in altered
wetlands and the success of replication efforts  at the site, but also
to gather data that can  be applied to other sites where wetland resto-
ration, alteration or replication is proposed.
  The information and proposed mitigative measures presented
in the wetland assessment enabled the U.S. EPA to develop its
recommendations  for design changes to the selected remedy,
including relocation of the  sedimentation basin.
CONCLUSION
  This wetland assessment procedure examines ecosystem, wild-
life habitat, hydrologic, water quality and socioeconomic functions.
Unlike other similar procedures, it also determines current impacts
and evaluates the impacts of proposed remedial actions. Based on
this information, the U.S. EPA can modify proposed remedial
actions in order to avoid or minimize adverse impacts. In situa-
tions  where impacts have already occurred or  are unavoidable,
appropriate mitigative measures, such as wetland replication or
restoration, can be developed.
REFERENCES

1. Reed, Jr., P. B. "1986 Wetland Plant List—Northeast Region." National
   Wetlands   Inventory/U.S.   Fish   and   Wildlife  Service.
   WELUT-86/W 13.01, May 1986.
2. Cowardin,  L. M., Carter, V., Oolef, F.C. and LaRoe, E.T.. "Classi-

-------
  fication of Wetlands and Deep Water Habitats of the United States."          Assessment, Volumes 1 and 2. Center for Natural Areas, Gardiner,
  U.S. Fish and Wildlife Service. FWS/OBS/-79/31,  1979.                    ME „ Federal Highway Administration Report Nos. FHWA-IP-82-23
3. Adamus, P.R., and Stockwell, L.T. "A Method for Wetland Functional          and FHWA-IP-82-84, March 1983.
                                                                    WATERWAYS AND WETLANDS RECLAMATION    435

-------
                                             Degree of  Cleanup:
                          SARA  121(d),  ARARs and Mining  Sites
                                                     Don  C. Porter
                                                    William E. Cobb
                                                      CH2M HILL
                                                       Denver, CO
ABSTRACT
  Section  121(d) of SARA addressed the issue of cleanup levels
at Superfund sites by requiring the analysis of federal and state
applicable or relevant and appropriate standards, requirements,
criteria or limitations  (ARARs).  The U.S. EPA has set three
categories of ARARs:  chemical-,  action- and location-specific.
  Mining sites, by virtue of size, volume and type of waste and
the  potential extent of cross-media contamination, pose a chal-
lenging task when analyzing ARARs to assess the degree of cleanup
required at each site. Only by clearly identifying the waste types
and media interaction can the ambient or chemical-specific ARARs
be identified. However, waivers may be required because of the
technical infeasibility of treating the enriched chemical conditions
of these sites. Location-specific ARARs often raise questions of
protection of heritage versus protection of environment whenever
mine camp or boom town vestiges of the past are threatened by
the  cleanups of the present. Finally, the current  development of
regulatory requirements (RCRA Subpart D) for cleanup at mine
sites are still under development, causing difficulties in determining
some action-specific ARARs.

INTRODUCTION
  The remediation of Superfund sites invariably raises the question
"How Clean is Clean?" Congress addressed this issue in SARA
by the addition to CERCLA of Section 121(d) entitled "Degree
of Cleanup." Simply speaking, Congress codified U.S. EPA policy
"CERCLA Compliance with Other Environmental Statutes." The
basic tenet of the policy, now law, is that in the determination of
the  degree of cleanup for a site, the initial consideration for
protection of human health and the environment should rely on
existing legally applicable or relevant and appropriate standard,
requirement, criteria, or limitations (ARARs). An important aspect
of planning for site remediation  now becomes  the selection of
ARARs.
  The U.S. EPA is in the process of developing a rational approach
in the  selection of ARARs1.  To this end, three categories of
ARARs have been specified:  (1) chemical-specific,  (2) action-
specific and (3) location-specific. ARARs also come in two types:
federal and state.
  Chemical-specific ARARs establish health- or risk-based  con-
centration limits in various environmental media for  particular
hazardous substances.  These ARARs typically will drive the  level
of cleanup by establishing a floor of protectiveness for certain
regulated contaminants. Location-specific requirements set restric-
tions on certain types of activities such as those in wetlands, flood
plains and historic sites.  Location-specific  ARARs generally can

436    MINING WASTES
have more effect on remediation cost than on the degree of cleanup.
  Action-specific ARARs are technology-based restrictions that
are triggered by the  type of cleanup action under consideration
at the site and generally do not resolve the issue of the appropriate
level of cleanup.

MINE  SITES
  ARARs are not intended as national cleanup standards, and each
site should be approached as a new and unique situation. Mining
sites, because of their large size (ten to several hundred square miles)
and diverse wastes represent a challenge in identifying ARARs.
These sites can have the following waste types:

  Tailings impoundments
  Slag  piles
  Unconsolidated waste piles
  Contaminated soils
  Acid drainage

  These waste  types have similar characteristics, primarily being
inorganic-rich  and consisting of large voliimes. However,  the
specific chemical constituents of these waste types are extremely
variable.  Individual tailings impoundments, for example, can be
chemically heterogeneous. That is, the chemistry of the tailings
could vary according to the initial ore type, ore grade, metals prices
and metallurgical process used at the time of the mining activity.
  Affecte"d media at these sites might include air from inorganic
laden dust, surface water from contaminated run-off, groundwater
from  tailings  impoundment seepage and soils from smelter
emissions or erosion of waste piles. Considerable cross-media con-
tamination might also exist, e.g., surface water run-off from waste
piles seeping into groundwater beneath a stream bed (Figure 1).
The volume of  waste, the extent of cross-media contamination and
the sheer area covered by those sites can be intimidating when at-
tempting to establish the degree of cleanup in accordance with the
mandate  of Congress.

Chemical-Specific ARARs
  In order to comply with the overall goal of Section 121(d), which
is protection of human health and the environment, the approach
toward specifying chemical-specific ARARs should be modelled
directly along the lines of the approach of establishing endanger-
ment or risk assessments for any  NPL site. That  is, one must
establish the relationship between the waste types and characteris-
tics and the affected media. This relationship narrows the universe
of laws and regulations to a medium-specific and thus chemical-
specific basis. This greatly aids in the determination of the degree

-------
                AIR
SURFACE
SOLIDS AND
ACID DRAINAGE
*
GW




ONSITE
SURFACE
WATER
                       OFF-SITE SW
                           Figure 1
                       Media Interaction
of cleanup for the site.
  Any given mining site typically will contain one or more of the
waste types previously listed with subsequent media interactions
illustrated by Figure  1. By identifying the waste types and affected
media, the process of choosing chemical-specific ARARs can be
simplified. Should a site have all of the possible mine waste types,
as many do, then the possible chemical-specific ARARs could be
those listed in Table 1.
                          Table 1
                   Chemical-Specific ARARs


 MEDIA              ARAR


 Air                Clean Air Acti

                      National Ambient Air Quality Standards
 Ground Hater       Clean Hater Act:

                      Maximum Contaminant Levels  (HCL)

                      Maximum Contaminant Level Goals (HCLG)

                    State Aquifer - Specific Standards



 Surface Hater       Clean Hater Act:  HCL, MCLG


                    Safe Drinking Hater Act:

                      Hater Quality Standards

                        State Stream Standards

                        State Non-Degradation Standards

                        State Agricultural Use Standards
  While  that  rational approach  to  establish  chemical-specific
ARARs is a tremendous aid, other conflicts exist at mine sites more
than other CERCLA sites. Mining sites, by virtue of the geologic
phenomenon that created the economic riches, possess characteris-
tics that can create a nightmare when establishing chemical-specific
ARARs.  The  naturally-occurring geochemical conditions  often
make the achievement of ARARs, particularly conservative
standards for  protection of sensitive biota, virtually impossible.
A good example of this conflict would be the Water Quality Criteria
(WQC) for protection of aquatic life (chronic criteria) for zinc,
currently set at 86 /*g/l adjusted to a hardness of 100 /ปg/l. Back-
ground levels or naturally occurring levels of zinc at the Red Dog
Deposit in the Brooks Range of Alaska (a sulfide ore body not
yet affected by man's activities) occur at levels up to hundreds of
thousands /tg/1 in surface waters of the area2. This situation could
be repeated at sites in the western United States located near similar
types of ore bodies. Other metals, including cadmium and lead,
that  are set at levels  to  protect sensitive biota, also  occur
"naturally" at much higher levels.
  SARA, section 121(d)(4), provides for a waiver should compli-
ance with ARARs prove to be technically impractical. Geochemi-
cal issues, such as geology, chemical specification and deposition
of the contaminant (i.e.  aerosol vs. erosion) need to be evaluated
in the context of technical feasibility of meeting the ARAR. Should
the geochemical characterization of the site demonstrate the tech-
nical impracticability of meeting a chemical-specific ARAR, then
a waiver may be necessary. However, provisions made in lieu of
the particular ARAR that is waived must still be  protective of
human health and the environment.

Location-Specific ARARs
  Common location-specific ARARs at most NPL sites include
flood, plains protection and wetlands enhancement. These two
ARARs can also be important at mine sites  that are located near
major streams or rivers. These same mine sites typically are located
in mountainous regions, where there is a rich history of early settlers
and boom towns. The historical aspects of a mining site often create
conflicts between laws meant to preserve our nation's heritage and
laws meant to protect our nation's environment.  Historical-related
laws that must be examined for mine sites  include:

•  National Historic Preservation Act
•  Archaeological Historic Preservation Act
•  Historic  Sites Act

   These acts quite rightfully protect history. Yet the cleanup of
mine sites might often entail moving literally mountains of material,
thereby imperiling abandoned mine camps or mill sites that date
back to the 1800s.

Action-Specific ARARs
   Finally, action-specific ARARs are difficult to decipher at mine
sites due to the current development of regulatory requirements,
specifically, the application of RCRA to mine sites3. Currently,
all wastes from the extraction, beneficiation and processing of ores
and minerals are conditionally excluded by RCRA under the Bevill
Amendment from regulation as hazardous  waste under Subtitle
C of RCRA.  Jurisdiction over mine waste currently falls within
Subtitle D of RCRA. The U.S. EPA, under Section 8002 of RCRA,
must submit reports to Congress on mine wastes and then deter-
mine which of the wastes should be regulated under Subtitle C.
   Meanwhile, the regulations  under Subtitle D must be drafted.
This is currently underway. The Mining Waste Regulatory Develop-
ment Group has been meeting since January 1986 to develop the
Subtitle  D regulations.
   The relevancy of other mining related  laws and  regulations is
also at issue in assessing action-specific ARARs. These laws include:

   Mining Law of 1872
   Organic Administration Act of 1897
   Mineral  Leasing  Act of 1920
   Federal Land Policy  and Mineral Act of 1976
   Surface Mining Control and Reclamation Act of 1977
   Uranium Mill Tailings Radiation Control Act of 1978

   Labor safety regulations under OSHA  or  MSHA also may
require scrutiny. Additional action-specific ARARs, particularly
state requirements, may involve air quality monitoring, monitor
well construction, solid waste  disposal, NPDES requirements for
                                                                                                    MINING WASTES    437

-------
water treatment and others.

CONCLUSION
  A rational approach is available to tackle the question of "How
Clean is Clean?" at mining sites in a manner that meets the intent
and specifications of Congress. The technical feasibility of meeting
chemical-specific ARARs, particularly surface water and ground-
water standards, must be addressed on a site-by-site basis. Waivers
from certain chemical-specific ARARs may need to be sought under
the provisions written into SARA. The location-specific ARARs
may result in conflicts between local/national heritage and environ-
mental cleanup. Meanwhile, critical regulatory issues concerning
action-specific ARARs are still in development. In summary, con-
flict in the selection of ARARs at mining sites exists. A clarifica-
tion of policy balancing the nation's goals and resources may be
needed to resolve the conflicts of SARA 121(d), ARARs and mining
sites.

REFERENCES
I.  U.S. EPA, "Interim Guidance on Compliance with Applicable or Rele-
   vant and Appropriate Requirements," OSWER Directive 9234.0-05.
   July 9. 1987.
2.  U.S.  EPA,  "Environmental  Impact Statement—Red Dog  Mine
   Project." 1984.
3.  U.S. EPA, "RCRA Subtitle D Regulatory Program for the Manage-
   ment  of  Mining  Wastes," Draft Regulatory  Development Plan.
   Mar. 3. 1987.
438    MINING WASTES

-------
          Clearing and  Abandonment  of Deep Water Wells  at  the
                                     Tar Creek  Superfund Site
                                                  Michael Howar
                                           Engineering Enterprises Inc.
                                                  Long Beach, CA
                                                  Douglas Hayes
                                           Engineering Enterprises Inc.
                                                  Long Beach, CA
                                                    Bart Gaskill
                                     Oklahoma Dept. of Pollution Control
                                                Oklahoma City,  OK
ABSTRACT
  The  Tri-State Mining  Area is  located at the junction of
Oklahoma, Kansas and Missouri. The Tar Creek Superfund area
includes the Picher Mining District located within the Tri-State
Mining Area in northeastern Oklahoma and southeastern Missouri.
The area was mined for metallic zinc  (sphalerite Zns) and lead
(galena PbS). In the mid 1930s, the smaller mining operations were
centralized into a larger milling operation.  The closure of the
smaller mills resulted in the abandonment of water wells screened
in the Roubidoux Formation associated with the smaller mills. In
the late 1960s, the Picher Field mining District was completely
abandoned when the recoverable ore was depleted.
  In March 1981, as part of a groundwater investigation in Ottawa
County, Oklahoma, The United States Geological Survey (USGS)
conducted a study of several wells in the Roubidoux Formation.
A spinner log indicated downward flow in the Consolidated No.
2 Well. In late  1982,  the Oklahoma Water Resources Board
(OWRB) became involved in a water well sampling program to
study the effect of acid mine water on the Roubidoux Formation.
  Surface waters enter the mines through inflow structure such
as shafts, collapse structures, etc. After heavy rains, acid mine water
is discharged into the underlying Roubidoux  aquifer via the
abandoned water wells and test borings.
  Deep water wells and test borings near Picher, Oklahoma pro-
vided preferred vertical pathways of acid mine water contamination
to the Roubidoux aquifer from past underground mining opera-
tions of the overlying Boone Formation. The Roubidoux aquifer
is a principal source of water supply for industrial, municipal and
agricultural purposes covering the areas of Northeast Oklahoma
and Southeast Kansas. This paper  details  the  plugging and
abandonment procedures for the wells, the procedures and difficul-
ties involved in locating and clearing these wells and discusses the
numerous regulatory requirements and concerns.

INTRODUCTION
  The Tar Creek Superfund  site is located within the Tri-State
Mining Area in Northeastern Oklahoma, Southeastern Kansas and
Southwestern Missouri (Figure 1). The principal ores mined in the
area were sphalerite  and  galena. The  principal ore veins were
located within the Boone Formation which is Mississippian in age
and generally consists of dense, massive dolomites and chert layers.
  Economic deposits of sphalerite and galena were first discovered
in Southwestern Missouri in the area around Joplin, Missouri, in
the late 1800s. These shallow deposits originally were mined from
slopes and surfaces pits. As time progressed, the shallow deposits
were mined out, and mining operations began operating out of
increasingly deeper pits and shafts as the ore body progressed and
dipped to the west and northwest through Northeastern Oklahoma
and Southeastern Kansas. During the early to mid-1900s, exten-
sive mining operations were developed in the Kansas and Oklahoma
portions of the field in order to supply materials needed for the
war effort. The lack of adequate high quality surface water supplies
caused the mining and milling operators to obtain their process
and wash water from the underlying Roubidoux aquifer.
  The Roubidoux aquifer is composed of massive limestones, dolo-
mites and cherts with minor discontinuous shales. In order to obtain
the necessary wash water, numerous deep water supply wells were
completed  within the Roubidoux  and underlying Gasconade
Formations using cable tool methods. In most cases, these wells
were constructed by installing a surface casing to a point below
the Boone Formation and occasionally into the underlying Cotter
           KANSAS     Piiซปbซrj;
                 Panoni
MISSOURI
         Springfield
    BcrtUtvlItt
                                       ARKANSAS
                         Figure 1
          Site Location Map of Tar Creek Superfund Site
                                                                                             MINING WASTES    439

-------
and Jefferson City dolomites. From this point, the wells were com-
pleted using open borehole methods to their total depths, generally
between 1,000 to  1,200 ft below land surface.
  During the 1930s, many of the smaller mills were closed in favor
of larger more efficient modern centralized mills. In most cases,
the former water supply wells were either abandoned and left open
after removing the pumps or were plugged using mill tailings as
fill material. In the late 1960s, the remainder of the mining and
milling operations were closed due to economic reasons and the
reduction in domestic mining of materials. With the advent of the
termination of mining operations, the shafts and drifts subsequently
filled with acidic groundwater. The low pH of the groundwater
within the shafts and drifts was due to the chemical degradation
of the remaining sulfide minerals present within the Boone  For-
mation. The acidic nature of this groundwater quickly corroded
the casing of wells completed through the Boone Formation and,
due to differences in the potentiometric levels of the two aquifers,
moved downward through the open or gravel filled bore holes and
entered the Roubidoux aquifer. Since the Roubidoux aquifer is clas-
sified as a sole underground source of drinking water  (USDW)
aquifer for the Tri-State area, protective measures had to be un-
dertaken through the State and Federal Superfund programs to
prevent  the downward flow of contaminated groundwater from
the Boone Formation to  the Roubidoux  Formation.

Regulatory Considerations
   In November 1979, surface discharge of acid mine water was
discovered near Commerce, Oklahoma. A sample of the orange-
colored water was obtained and delivered to the Oklahoma Water
Resources Board. Since that date, a continuous discharge of highly
mineralized acid mine water from the flooded underground lead-
zinc mines of the Richer Field in Ottawa County, Oklahoma, has
occurred. In June 1980, the Governor of Oklahoma formed the
Tar Creek Task Force, which was comprised of representatives
from 24 local, state and  federal  agencies. This Task Force was
responsible for implementing field studies in the Tar Creek area,
determining the most practical and  cost-effective solution to the
problem and  also was responsible for oversight of the remedial
action. At the time the Task Force was created, the Governor also
designated the Oklahoma Water Resources Board (OWRB) as the
lead agency in studying the pollution problems in the Tar Creek
area and named the OWRB's Water Quality Division Chief as the
technical co-chairman of the Task Force. The OWRB has continued
to lead all studies, research and construction supervision of the
various  activities that have taken place.
  In November 1981, the U.S. EPA published the first NPL. The
Tar Creek site near  Richer,  Oklahoma was subsequently ranked
within the top 10 on the NPL and was, therefore, eligible to receive
funding under CERCLA. The State of Oklahoma then proceeded
to enter into a cooperative agreement with the U.S. EPA to under-
take the RI/FS for  the Tar Creek area.
  In June 1982, the OWRB, through the Tar Creek Task Force,
studied various aspects of the Tar Creek area. These studies  were
ongoing through the end of 1983, and in January 1984 the Task
Force  submitted  to the Governor and the  U.S.   EPA its
recommended remedial plans for mitigating acid mine water con-
tamination in Northeastern Oklahoma. Included in the Task
Force's  strategy was the plugging of abandoned water wells and
exploration holes. It was discovered that the deep underground
boreholes could serve as pathways for downward migration of the
contaminated water from the overlying Boone Formation to the
underlying Roubidoux aquifer.
   The U.S. EPA  and the State of Oklahoma's first priority was
to protect the quality of water in the Roubidoux Formation. This
fresh water aquifer currently serves Northeastern Oklahoma,
Southwestern Kansas, Southwestern Missouri and Northwestern
Arkansas  and the degradation of such an aquifer would have a
severe impact on this four state region. The first step facing the
Task Force was to locate all deep wells and exploration holes that
 had the potential  to contaminate  the Roubidoux  aquifer. The
 preliminary information which was  obtained from former miners,
 land owners and local residents had to be backed by supporting
 documents. The following sources were used to compile a reliable
 data base:

 •  Mine Maps — The United States Bureau of Mines had conducted
   extensive research on the location of the exploration holes and
   the  extent of underground mine workings in the Tri-State
   mining district. Microfilm copies of information were obtained
   from the Oklahoma Geological Survey and Kansas Geological
   Survey for  their respective state's portion of the Tri-State
   mining area. Additional maps of the area were provided from
   the Spiva Library at Missouri Southern State College in Joplin,
 Missouri.
 •  Mining Companies —  Eagle-Picher Research Laboratory in
   Miami, Oklahoma provided more specific data on the deep wells
   in the Oklahoma portion of the Richer  Field. Although the data
   were not complete, they did provide a valuable basis for detailed
   information including lithologic features, physical features such
   as well and casing depths and any additional plugging data for
   the few wells which had been properly plugged by the company.
 •  Government Agencies — During the  data base development,
   several government  agencies  were contacted. The principal
   agencies that provided the most  substantial information were
   the U.S. GS, State Geological Surveys of Oklahoma, Kansas and
   Missouri; and the Groundwater  Division of the OWRB.
 •  Publications — During the course  of mining activities, there were
   many reports published  by various engineers and/or geologists
   working in  the Tri-State mining district. A few  reports were
   found to have relevant information on locating deep wells and
   exploration  borings in the mining  area.
  Another area of concern to the  U.S.  EPA and the State of
Oklahoma, was the possibility of adverse  effects of well plugging
activities on the environmental resources in the area. The resources
are in general classified as air, water, noise, biological, cultural
and socioeconomic environments. The well plugging project caused
no changes in  the ambient air quality or noise levels, except on
a temporary basis and at minimal levels during the actual plugging
period. The project required no displacement of existing residences
or disturbed residential area. The project implementation involved
no agriculture, park, public lands or historical sites; therefore, no
cultural and socioeconomic impacts  were incurred.  The actual
plugging operations generated only an insignificant volume of
drilling circulation water which was stored in a reserve pit and
properly disposed of when plugging operations were completed.
Therefore, no adverse impacts on the water resources and biological
environment of the project occurred  as a result of the well plugging
project.

Regional Geology
  The  major  geologic  and tectonic features of  the Richer,
Oklahoma mining area include: parts of  the Cherokee platform-
Ozark  uplift, the  Bourbon Arch, Nemaha Ridge and Arkoma
Basin.
  The  rock formations exposed at  the surface of the ming area
include: Mississippian and Pennsylvanian rock units with a gently
northeastward  dip  of  approximately 80 ft/mile.  The  Boone
Formation is the principal host for the ore deposits in the area.
It consists mainly of fossiliferous limestone and thick beds of nodu-
lar chert. This formation, which is approximately 400 ft thick, can
be further subdivided into seven members: St. Joe Limestone,
Reeds  Spring,  Grand Falls Chert,  Joplin, Short  Creek Ookite,
Baxter Springs and the Moccasin Bend.
  The  Boone Formation is underlain by  Cambrian and Ordovi-
cian formations composed primarily of dolomite  and chert with
some  sandstone and minor shales.  Of these formations, the
Roubidoux, Jefferson  City,  Dolomite,  Cotter  Dolomite and
Chattanooga Shale are significant to the purpose of the Tar Creek
440     MINING WASTES

-------
project (Figure 2).
SyMfli
I
MISSISSIPPIAN
MISSISSIPPIAN
AND DEVONIAN
OflDOVKIAN
CAMBRIAN
Sซcป.
I
!
1
ซ
KinOfhookian and
Upper Devonian

Lowซr
Ordovician
Uppw
Ca/nbnan
PRECAMBRIAN
Group. Formation
or rrwmtef
KMtwQrm*
ftuซซUI SwvfcttTM Uw<ปw
(o* Boon FormMon)
StvwwMTt SftftM
Oonปl*y LxnMlofW U*mtttr
|R*ซI vtf otMra. 1K5)
Wviw S*ndซonซ UซmMr
H^™.*,™,
Ftrw*** Siuto
BMMV* SVKtMTM
MndrrfU UmMlon*

i
MoccumBซM
Mซmbซr
BuwSpnnfft
Mwnbw
Sfton Crack OoUซ MtmOif
JocWiMwntwr
Grand FiM
Qwrt MOT**
RMMSprmg
Mซmbซr
SLJo* LimMlorw
Mwntwr
CftMwwoea Sfwto
CttMr DotonM*
JvflOTon C*y DotofM*
FkwtXkxn FonriMwri
GMOMWI* OotomiH
VM Bw*n Formrton
ftw>* Dotomtt
D>vn FomMflon
Bonr^Xfft OotormU
LMnon* SWNMWW
GrwwaM
votaoo
B*d
(Infernal Lซflซr Ctaaafflcation)

•
C
0
E
F
Q
H
J
K
L


H
Q
P
0
R

                           Figure 2
   Regional Stratographic Cross-Section of Tar Creek Superfund Site
  The Boone Formation is overlain by the Quapaw Limestone in
some parts of the main Picher district. The Chesterian Series
represented by the Hindville Limestone, Batesville Sandstone and
Fayetteville Shale, generally forms a disconformable contact with
the Boone Formation and the Quapaw Limestone. Both the Hinds-
ville and the Batesville are locally mineralized, especially in the
eastern portion of the Picher mining district.

Hydrology  of the Mined Area
  Tar Creek and its main tributary, Lytle Creek, are the two major
streams that flow through the main part of the Picher Field. The
movement of water down these tributaries has been modified by
mill-tailings piles  adjacent  to and sometimes within the main
channels. In some areas, diversion dikes were constructed to keep
water from overflowing into shafts. At the time of mining, surface
water was generally of poor quality and was not used for domestic
consumption.
  Groundwater is the principal source of water for domestic and
industrial users adjacent to and within the Picher Mining District.
   The Roubidoux and Boone Formations are the principal ground-
water aquifers in this region. All of the public water supplies and
most of the industrial supplies in Ottawa County come from wells
drilled into the Roubidoux. The Roubidoux Formation is a 160-ft
thick sequence of Ordovician cherty dolomite interbedded with thin
sandstone beds.  This aquifer is generally between 900 to 1,000 ft
deep in the mining area (Figure 3). By the 1920s, a number of mill
operators drilled wells into the Roubidoux to augment their water
supplies. By 1947, the potentiometric surface of the Roubidoux
was substantially lower. Pump lifts were as much as 500 ft. Water-
level data obtained from the city of Miami suggest that the decline
has  stabilized since 1975. The potentiometric surface  of the
Roubidoux near Miami appears to have remained about 320 ft
above mean sea level  since 1975.
                                                                                            Figure 3
                                                                     Generalized- Regional Cross-Section of Tar Creek Superfund Site
  Next to the Roubidoux Formation, the Boone Formation is the
most important source "of groundwater in Ottawa County. The
groundwater in the Boone has been utilized only to a small extent
because of the variability in water yields. Contaminated surface
waters may enter the Boone readily through fractures and/or sink-
holes. The possible condition is a serious drawback to the utiliza-
tion of groundwater from the Boone for a reliable public supply.
  At present, recharge to the mines comes from natural infiltra-
tion through fractures and solution cavities as well as from inflow
to abandoned shafts, boreholes and  collapse features. Water
samples taken from  mine shafts from 1975 to 1977 indicated the
presence of metals such as cadmium, lead, zinc, iron, manganese
and nickel. Water levels continued  to rise in the mine workings
until the workings became completely full of water. Acid water
charged with above-normal concentrations of iron, lead and zinc
began to discharge to the surface through abandoned boreholes
in late 1979. Two main discharge points occur south of the brick
plant in south Commerce and near the confluence of Lytle and
Tar Creeks.
  Considerable interest and concern  exist about potential contami-
nation of surface-water supplies from mine water discharges. Also,
there may be a hydraulic connection between the Boone Forma-
tion and the underlying Roubidoux Formation, downward migra-
tion of contaminated mine water into the Roubidoux may occur.
Furthermore, mine water also may  migrate laterally through the
Boone and possibly contaminate nearby rural, domestic and stock
wells,

Subsidence Potential
  In the Picher Field, beds  were mineralized over  such a large
vertical extent that  some mining chambers with ceiling heights
greater than 90 ft were created during active mining. The over-
                                                                                                   MINING WASTES    441

-------
lying rock strata, principally Pennsylvanian shale and sandstone,
are not very competent and further stripping of  pillars before
abandonment of the area led, in some places, to roof instability
and subsequent subsidence. Since the abandonment of the Richer
mining area, many potential safety hazards and environmental
concerns have been recognized.
  Because of the shallowness of the deposits, the relative ease with
which  they could be mined and milled and the wide technical
experience gained from earlier operations in adjacent subdistricts
of the Tri-State region, numerous small but efficient mining com-
panies were organized. In the main part of the field, an almost
continuous underground network of mine workings extended from
near Eagle-Picher's central mill northward  into Kansas.
  Open shafts and surface-collapse features associated  with the
abandoned underground mine workings probably presented the
greatest visible hazard potential in the Richer  Field.

Site Preparation
  The roads throughout the project area allowed ready access to
most of the well sites  in the region. Site preparation, which was
performed as needed,  consisted mainly of clearing  the vegetation
and debris from around the well area, constructing well pads and
excavating mud pits. The mid pits used in drilling were lined with
Visqueen  which allowed circulated water to be contained within
them.
  An initial site survey and well sounding was conducted to deter-
mine  the  amount of debris which had to  be moved  prior to
commencing well clearing and plugging operation. Action was
taken on a case by case basis depending on this field observation.
Well Clearing and Plugging Methods
  In 1983, a feasibility study was conducted at two abandoned wells
in the vicinity of Richer, Oklahoma. The purpose of the study was
to develop methods which could be used and expanded upon to
safely and successfully plug the abandoned  wells. The feasibility
study indicated that clearing and plugging the wells was technical-
ly feasible using a mixture of water well and oil well drilling tech-
nologies.
  The main objective of the Tar Creek project was to plug and
abandon deep water wells which wee thought to provide vertical
pathways  of acid mine  water  contamination to the Roubidoux
aquifer from past underground mining operations in the Boone
Formation.
  The project was approached in three ways which theoretically
would achieve  the  same end  result, to inhibit further  vertical
migration of contaminants:

• The first scenario would include the clearing and plugging of
  the wells to the Roubidoux  aquifer. This would  occur  if the
  total depth of the well could be reached in an economically
  feasible length of time.
• The second scenario  would include those wells which could not be
  cleared  of debris in a reasonable time-frame, therefore making
  it prohibitively expensive to continue further operations. In these
  instances, an alternative plan of action would include the clearing
  of the wells through the base of the Boone Formation across
  the Chattanooga shale (Figure 3). This shale unit is thought to
  be extensive in the area and relatively impermeable, thus creating
  a natural barrier against further downward migration of con-
  taminants.
• The third scenario includes all remaining wells which could not
  be cleared through the Chattanooga Shale unit. These wells were
  handled on a case by case basis with the overall  result being to
  attain the maximum possible depth before it became uneconomi-
  cal to complete the well. At that point, the well would be plugged
  from total depth to surface.

  The drilling equipment used  to open the boreholes consisted of
medium-sized water well drilling equipment with  rated working
depths of 1,300 to  1,500 ft. During the FS, only minor lost cir-
culation problems were encountered. The first phase of the clearing

 442    MINING WASTES
and plugging, which was supervised by IT Corporation and con-
ducted by Hillard Drilling, indicated that downhole conditions were
considerably more complicated than originally anticipated. As a
result of the different downhole condition, additional specialized
equipment had to be utilized to reach the necessary objectives. The
necessary equipment consisted mostly of oil field tools for fishing
operations, washover pipe for regaining circulation in wells after
drilling through lost circulation zones, mine workings and drifts
and various types of mills, magnets and downhole assemblies which
are used for drilling casings and debris which cannot be otherwise
removed from the borehole. The additional tools were added to
the contract during Phase I  by means of a change order.
  During the Phase 2 clearing and plugging operations, which were
supervised  by  Engineering Enterprises,  Inc. and conducted by
Williams Water Well, the same general methods and equipment
were utilized. One change order was processed to add some neces-
sary  additional tools, bits and materials not listed in the  original
contract.
  The basic procedure used to clear the wells involved the circula-
tion  of drilling  fluids down  the hole  to remove fill and cuttings
from the boreholes. When lost circulation zones were encountered
(ie.,  mine workings, caverns, or washouts), a string of washover
pipe was lowered into the borehole to seal off the zone and allow
continued penetration.  In numerous wells, junk iron, in the form
of iron pipe or metal fragments and cable, were encountered. In
these wells, a series of mills was used  to either drill  up and circu-
late the material to the surface or "dress-up" the top of the material
so that  it could be successfully "fished"  or removed from  the
borehole.
  Fishing operations consisted of running either casing spears,
cable spears, overshots, box  taps or taper taps into  the boreholes
to grab the material at the bottom of the borehole. Fishing opera-
tions continued in this manner until: (1) nothing could be removed
from the hole, (2) all material was removed from the borehole or
(3) it was determined that the objective could not be reached within
the budget.
  Washover shoes were successfully used to straighten or  mill  out
collapsed and broken  casings at  several well sites.
   In relatively few instances, the clearing procedure only included
the time it  took to trip the drilling pipe down to the bottom of
the well. This, however, was  not the normal condition in the area.
An average well took  approximately  10 days to close. This time
span was mainly due to the rocks, dirt, iron pipe, railroad spikes,
wood, cable and other debris used as fill material when  the well
was  abandoned earlier. In the span of only 10  ft, the driller may
have had to change to a different tool four  times, thus  making
the job of tripping pipe in and out of the drill hole an all day event.
Many times the well would span a mine opening. If the casing was
deteriorated, then  hope of dropping through the roof of the mine
and  finding the hole again at the floor of the mine was  limited.

Geophysical Logging
  After each well was cleared to the maximum  obtainable depth,
the well bore was geophysically logged by the U.S.  OS. The logs
which were taken indicated the presence and location of the Chat-
tanooga shale  where present along with the well bore diameter,
depth and  type of the  formation behind the casing in the well.
  Several types of borehole geophysical logs  as well as video logging
were used during the project. The information generally sought
after consisted  of:

•  Location of the top of the Roubidoux aquifer
•  Location of the Chattanooga  Shale
•  Condition of the existing casing
•  Size of the well bore
•  Total depth of the  casing
•  Presence or absence of cement in the annulus between the casing
   and the drill hole, if the casing extends below the Chattanooga
   Shale

-------
  The geophysical logs most commonly used were as follows:

  Caliper
  Natural gamma ray
  Gamma-gamma
  Neutron
  Spontaneous potential
  Single point resistance
  Normal resistivity logs
  Black and white television

  Each log and its usefulness for providing information necessary
for the plugging operation is discussed below.

• Caliper Log: The caliper log determines the average diameter
  of the borehole which  is needed to calculate the volume of
  cement required to fill it. It also was found to provide limited
  information on the integrity of the casing. The casing informa-
  tion is limited in that small holes may not be detected by the
  caliper arms and, if the holes  are open to mine workings, large
  volumes of cement could be  lost.
• Natural Gamma Ray Log: The natural gamma ray log in con-
  junction with other logs was used to determine the lithological
  sequence penetrated by the well. This is important in that the
  presence and location of the Chattanooga Shale could be deter-
  mined and the cementing plan constructed to  insure a solid
  cement plug through this zone.
• Gamma-gamma Log: The gamma-gamma log was used in con-
  junction with the natural gamma and the neutron logs to deter-
  mine casing locations and lithologies of formations present.
• Neutron Log: The neutron log, when used in conjunction with
  the gamma ray log, is also used to determine rock types as well
  as to detect the presence or absence of mine workings outside
  the well casings. Areas  in which the casings penetrate mine
  workings had to be carefully  evaluated prior to attempting the
  placement of cement slurry.
• Spontaneous Potential Log: The spontaneous  potential log
  provided information on the rock types in the borehole. It is
  not as precise as the neutron  and gamma logs and also can be
  influenced by the invading mine waters.
• Single Point Resistivity Log: The  single  point resistivity log
  locates boundaries between formation changes. The single point
  logs obtained from the wells plugged in this program were not
  highly definitive; however, we  feel that they will be useful in most
  cases.
• Normal  Resistivity Logs: Normal resistivity   logs provide
  information on the quality of water contained in the formations.
  Two normal resistivity logs were run in each well with electrode
  spacings of 32 and 64 in. It was felt that the invasion of the fresh
  drilling water in pervious strata could be distinguished from the
  preceding invasion of the acid mine waters. Some limited indi-
  cations of the expected  responses were detected, but the value
  of the normal resistivity logs to  the plugging operation  is
  doubtful.
• Video Logs: Video logs allowed visual inspection of the bore-
  hole. Video surveys were conducted  at several wells. In general,
  they provided visual verification of the presence of invading mine
  water through leaks in the casings. The video logs were also very
  useful in determining the condition of the old casings, allowing
  measures to be taken to prevent loss of cement into the mine
  workings
PLUGGING PROCEDURES
  The borehole was then partially filled with sand up to a depth
of approximately 900 ft below ground level. This was done to pre-
vent filling the more porous layers in the Roubidoux Formations
with cement grout. Filling the Roubidoux Formations with grout
was not felt to be necessary when a sufficient amount of grout could
be installed in the borehole above the Roubidoux.
  The sand  fill and cement  grout were placed in the well bore
through either 2-7/8-in.  tubing or 2-3/8-in. tubing (tremie pipe)
from the bottom up. This method prevents the accidental bridging
of the material in the borehole  and in the cementing operation
assured that cement grout is  placed in continuous homogeneous
plugs. The cement used  in these procedures was a half and half
mixture of API Class H cement and API Class A with 50%
pozzolans. This mixture was used because the high sulfates found
in the mine workings are detrimental to regular cement. Cellophane
flakes along with a thixotropic additive also were mixed with the
cement slurry to prevent movement or loss into the  surrounding
strata.
  Each well was grouted in three stages due to the potential for
hydraulically locking the pipe in the borehole if larger stages were
attempted. At one location, KS-3A,  this still occurred; however,
the tremie pipe was completely filled and encased in cement, thereby
assuring a competent plug in  the bottom portion of the borehole.
  Cement baskets were used  in several holes with lost circulation
zones or mine workings  in the Boone Formation to prevent the
loss of cement grout into the mines and caverns. The cement baskets
were placed into the hole and were supported from beneath by a
length of 4-in. PVC pipe which was in turn supported by the lower
cement plug. The cement basket bridged off the borehole above
the lost circulation zone and allowed the placement of a continuous
plug from the bridge to  the  land surface.
  Upon the completion of plugging operations, an aluminum
marker and magnet were installed in the cement at the top of each
well. This benchmark allows the future identification and location
of the well site.
Conclusions
  Approximately one third of the water wells were cleared and
plugged to their total depth. In some instances, problems were
encountered which so increased expenditures  that clearing and
plugging the well down to the Roubidoux was not feasible. In these
circumstances, the Chattanooga Shale below the Boone Forma-
tion became the target clearing and plugging depth. The Chatta-
nooga  Shale is an  impermeable unit which inhibits downward
migration of the acid mine water.  This unit was detected  by the
use of Geophysical logs  such as  Caliper, long, short normal,
Neutron, Gamma, Gamma Gamma and fluid resistivity logs sup-
plied through  the U.S. GS. Forty-eight wells were cleared and
plugged to a depth at or  below the Chattanooga shale.
  Thirteen wells could not be cleared and plugged below the floor
of the mine room. There was no casing present through the mine
room,  and the remaining well bore could not be located  on the
mine floor despite many attempts. These wells were plugged by
pumping a cement slurry into the mine room  creating a cement
cap over the well bore on the mine floor. Several boreholes were
also determined to  be vent wells which did not extend below the
floor of the mine.
                                                                                                   MINING WASTES    443

-------
                Mobilization  and Transport  of Metals  from Mine
                                Tailings to  an  Alluvial  Aquifer
                                              William B. Mills,  P.E.
                                              Steven A.  Gherini, P.E.
                                                  Tetra Tech,  Inc.
                                               Lafayette, California
                                                    Gary  Bigham
                                                  Tetra Tech,  Inc.
                                               Bellevue,  Washington
ABSTRACT
  Mathematical models were developed to predict the mobiliza-
tion and transport of arsenic, cadmium,  copper, lead and zinc
through mine tailings and into an alluvial aquifer. Oxygen trans-
port into and through the tailings and oxidation of reduced sulfur
were first simulated to determine the rates of acidity production
and the downward movement of the oxidizing zone through the
tailings. Up to 40,000 years will be required for the oxidizing zone
to pass completely through the deeper tailings.
  As the acidity generated within the tailings is leached into the
underlying alluvial aquifer, it is rapidly neutralized by the high solid
phase alkalinity in the alluvium. Consequently, formation of an
enrichment zone of metals is predicted near the tailings-alluvium
interface.

INTRODUCTION
  Over the past century, mining operations in the United States
have generated large quantities of solid wastes. A number of these
former operations have been declared Superfund sites by the U.S.
EPA.  One such site  is the Anaconda Smelter in Montana, which
operated for nearly  100 years before closing  in 1981.
  The tailings ponds and other features of the Anaconda Smelter
Superfund site are shown in Figure 1. The mine tailings are located
in the Opportunity and Anaconda Ponds. The tailings were gener-
ated by mechanical processes that reduced the ore to day- and sand-
sized particles. Using flotation, the mineral-poor portions of the
ore were removed from the process stream and slurried to the
ponds. However, the tailings still contain elements such as arsenic,
cadmium, copper, lead and zinc.
  The tailings ponds were created over various periods, as shown
in Figure 2. Portions of the Opportunity Ponds were first formed
in 1915 and are, in general, older than the Anaconda Ponds. The
ponds were flooded with water for many years to minimize dust
entrainment into the atmosphere. In the late 1970s and early 1980s
the  practice of flooding was discontinued.
             1910  1920  1930  1940  1950  1960  1970  1980
                   1	I	I	I	I	i_

A- Ponds Jjjjjjj — '• — • — ' 	
B-Ponds j_____^jj/tซ<:

D2 - Ponds
Anaconda 1 jjg
Anaconda 2
1 Active Disposal CD Flooded
1B75
1981
•983
1984
1BB3
^mmmmm
1953
EZ2 Surface Exposed to Ai
                          Figure 1
             Tailings Ponds and Other Features of the
                Anaconda Smelter Superfund Site
                                                                                       Figure 2
                                                                          Approximate History of Disposal Ponds
                                                                                And Flooding Operations.
  The cross-sections of the Anaconda and Opportunity Ponds and
the present water levels beneath the ponds are shown in Figure 3.
A portion of the tailings in both ponds is situated in the saturated
zone. The depth of saturation in the Anaconda Ponds may be up
to  8 m, while only the lower few  meters of parts of the Oppor-
tunity Ponds are saturated.
  Within the tailings ponds, the top 2 to 3 m are presently oxidized
(see Figure 4). Within the oxidized zone, the pH of the interstitial
waters is approximately 2.5. Below the oxidized zone, the tailings
are still reduced, and the pH is between 6.0 to 6.5.
444    MINING WASTES

-------
          Alluvium
          Groundwalvr Ltvel (static Water Level)
        |  Tilling! Pond Boundary
                             Figure 3
       Tailings Ponds Profile Showing Current Groundwater Level.
             i%63^:oxiDi';^
             w?\s? ป-1{tv-iซivY^"/J^^^ป^^'W.*^*^K*.*^^^h1?^!.*''ป
    TAILINGS
   ALLUVIUM
                             Figure 4
          Oxidized and Reduced Zones in Tailings Ponds, and
                 Typical Relationship of the Tailings
                    To the Underlying Alluvium.

SCOPE AND OBJECTIVES
  The objectives of this research are as  follows:

• To explain why the oxidizing zone is 2 to 3 m in the Opportunity
  and Anaconda Ponds, regardless of when the ponds were created
  or the de-watering history

• To determine when the oxidizing zone will reach the bottom of
  the tailings ponds, and the concentrations of metals in the inter-
  stitial water at that time
• To estimate the duration of leaching  after the oxidized zone
  reaches the bottom of the ponds
• To determine the concentrations of metals in the enrichment zone
  beneath the ponds

GENERAL CONCEPTS
  A conceptualization of the major processes controlling the gener-
ation of the oxidized zone is shown in Figure 5. Oxygen enters the
tailings pond in the gaseous phase and dissolves in infiltrating water.
  Oxygen is consumed during the oxidation of reduced sulfer.
While several sequential reactions are involved in the process, the
overall reaction  is given by Singer and Stumm-3 as:
FeS2(s) + i| 02  + Z H20 - Fe(OH)3(s)  + 4H +
                                                          (1)
  Pyritic sulfur (FeS2) is oxidized and, in the process, consumes
oxygen and produces acidity (liberates the hydrogen ions), caus-
ing the pH to decrease.
                                                                         DIFFUSION
                                                                        :02 (Air)
                                              I INFILTRATION
                                              (Dissolved
                                              Oxygen in Water)
                                                                                             ^CONSUMPTION
                                                                                                    :(DEWATERING)
                             Figure 5
                Generation of Active Oxidizing Zone at
                      Surface of Tailing Ponds.

   The oxidation process shown schematically in Figure 6 can be
thought of as occurring in discrete steps. Oxygen-saturated water
(8 mg/1 of oxygen at the site) in contact with pyrite initiates oxi-
dation. Twelve mg/1  SOl~ are  formed  in  accordance with
Equation 1 above. The deoxygenated water is reaerated by atmos-
pheric oxygen, and the oxidation process again proceeds. Twelve
mg/1 SO4~  are again produced to make a total of 24 mg/1 in
solution. Over time, high concentrations of sulfate are produced
in the aqueous phase. Data collected at the site indicate that sul-
fate concentrations between 2000-3000 mg/1  are present in  the
tailings ponds  and in the groundwater.
                        8 rug/102 produces a 12mg/l SO.

                           Figure 6
           Conceptualization of Iron Pyrite Oxidation.
  To simulate the oxidation process, the following mass transport
equations are required:

• Aqueous phase oxygen
• Gaseous phase oxygen
• Total reduced sulfur

  By assuming aqueous and tailings interstitial gaseous phase
oxygen are in equilibrium in accordance with Henry's Law, the
following mass transport equation results:
                            -    U 0W Co
                                                                                                                              (2)
                                               5Xป
                                                                     = J
                                                                       5z

                                                                    where

                                                                      ^wn
                                                                       U   =
                                                                      T   =
                                                                       * nซ/
         6  D
         vw 1-'o
          concentration of oxygen in aqueous phase
          volumetric fraction of water
          volumetric fraction of air
          velocity of water
          diffusion coefficient of oxygen in water
          tortuosity factor for diffusion of oxygen in water
                                                                                                        MINING WASTES     445

-------
   Xs   = total reduced sulfur (mass fraction)
   @B   = bulk density of tailings
   To2  = stoichiometric oxygen-sulfur ratio
   TM  = tortuosity factor for diffusion of oxygen in air
   KH  = Henry's Law constant for oxygen

   For total reduced sulfur, the mass balance of solid phase sulfur
 is given by:
          r*
—    K V    ฐw-
-  - rv As —	ฃ
                                                         (3)
   where the yet undefined symbols are as follows:

 Qw = saturation concentration of dissolved oxygen
 K   = first order oxidation rate

   In Equation 3, it is assumed that the oxidation of pyritic sulfur
 is linearly  dependent  on  the dissolved oxygen  concentration
 (normalized to saturation).
   The two  simultaneous nonlinear partial differential equations
 have been solved using the Galerkin Finite  Element Method-2.
 Linear basis functions have been chosen for simplicity.

 PREDICTED DISSOLVED OXYGEN
 PROFILES IN TAILINGS
   The simulations were begun by assuming that the oxidizing zone
 was not initially present and that the oxygen content of the water
 in the tailings was negligible. The propagation of dissolved oxygen
 into the tailings under these conditions is shown in Figure 7a. The
 influx of oxygen is initially quite rapid, and dissolved oxygen is
 predicted to be present at a depth of more than 1 m in only 0.02
 years. At approximately t = 0.6 years, a quasi steady-state dissolved
 oxygen profile is attained.

                         DISSOLVED CKVOEN.mg/l
                          Figure 7a
     Predicted Dissolved Oxygen Profiles in Tailings Ponds for
    t =0.2, 0.06, and 0.6 Years, Assuming  No Dissolved Oxygen
Present at t = 0.

  The quasi steady-state profile is established as the rate of sulfur
oxidation approaches the rate of influx  of oxygen. The relative
change in the pyritic sulfur is small over the time period simulated
in Figure 7a (less than 0.2% change), so that the sink of oxygen
is essentially  time-independent.
   The predicted dissolved oxygen and pyritic sulfur profiles at
t = 3000 years are shown in Figure 7b. The pyritic sulfur has been
                                                          completely oxidized in the top 3 m, and dissolved oxygen extends
                                                          to a depth of about 6 m. Thus the "active" oxidizing zone is about
                                                          3 m thick and is located between 3 to 6  m below the surface of
                                                          the  tailings ponds.
                                                                                   DISSOLVED OXYGEN mg/l
                                                                                   "OOOO        10000

                                                                                   PYRTTC SULFUR ing kg
                                                                                   Figure 7b
                                                                Predicted Dissolved Oxygen and Pyritic Sulfur Profiles
                                                                           In Tailings at t = 3000 Yean.
                                                            The predicted movement of the oxidizing zone for the next 50,000
                                                         years is summarized in Table 1. The active oxidizing zone is pre-
                                                         dicted to reach the bottom of the Opportunity Ponds (approxi-
                                                         mately  12-13 m) in about 20.000 years and the  bottom of the
                                                         Anaconda Ponds (20-25 m) in approximately 50,000 years. The
                                                         "active" zone gradually tends to decrease in thickness from about
                                                         3  m at the top  to 2 m at the bottom, and the rate of  movement
                                                         of this zone decreases with depth. Once the zone reaches the bottom
                                                         of the ponds, the active oxidation will continue for approximately
                                                         3000 years in  the  Opportunity Ponds and 5000 years in  the
                                                         Anaconda Ponds.
                                                                                             Table 1
                                                                                    Movement of Oxidizing Zone

tine, yr
3000
10000
20000
30000
40000
SOOOO
location of center of
oxidizing zone, n
4.5
9.5
12. S
16
19
21
thickness
of zone. •
3.5-3 •
2.S-3 •
2.S-3 •
2.5 •
2.0 ซ
2.0 ซ
tine to MVC
1 •. yr
600
1000
1500
1800
2000
2400
                                                          MOBILIZATION AND TRANSPORT OF
                                                          METALS IN TAILINGS
                                                            Based on the concept of an active oxidizing zone that moves
                                                          through the tailings, the approach to predicting metal transport
                                                          is to focus on the generation and transport of metals in the active
                                                          zone. This is done in an attempt to simplify the analyses, while
                                                          at the same time addressing the essence of the problem.

                                                          Approach to Metal Generation in Oxidizing Zone
                                                            A mass balance for metals and arsenic in the active zone can
                                                          be expressed as follows:
446    MINING WASTES

-------
Aqueous Phase Balance:

<(Cra,w) + ^(XP,B) +  u^a = D^J- - -?|




Primary Mineral Phase Balance:
           —    km Xn
                                                         (8)
                                                         (9)
 where

   Cm  =
   XP  =
   XM  =
   'S   ~~
   P   =
   Dr  =
   km
   D
         aqueous phase solute concentration in oxidized zone
         adsorbed phase concentration
         mass of element in primary mineral/mass of rock
         solids density
         rate of precipitation of element in zone
         rate of dissolution of secondary mineral
         fraction of volume occupied by water
         fraction of volume occupied by solids
         specific mineral weathering rate
         dispersion coefficient
  Model predictions for solutes in the oxidized zone are summa-
rized in Table 2 and compared to A-pond lysimeter data. While
both results and predictions are quite variable, there is a general
agreement between  the two. The range of  model  prediction
represents a range of seepage velocities, solid phase concentrations
and partition coefficients.
                          Table 2
          Predicted and Measured Aqueous Phase Metal
          Concentration in Near Surface Oxidized Zone
 Solute    Predicted Concentration, mg/1   A-pond Lysimeter Data,  mg/1
                                                                   where

                                                                     Cmi = metal concentration in zone i due to primary weather-
                                                                           ing and adsorption
                                                                     A   = cross-sectional area through which metal M(z) is passing
                                                                     6wi = volume fraction occupied by water in zone i
                                                                     Uj   = seepage velocity in zone i
                                                                     ATj = time interval over which leaching in zone i is  active

                                                                     For each depth interval, the concentrations shown are advected
                                                                   out of the oxidized zone as the active zone moves downward into
                                                                   the tailings. Most of the weathering of metals is predicted to occur
                                                                   in the top 10 m of the tailings ponds. In the zone nearest the sur-
                                                                   face,  the upper end of the concentration range for copper, zinc
                                                                   and cadmium will not actually persist over the entire time interval
                                                                   shown.  Complete leaching could  occur before then.  Below that
                                                                   depth, predicted weathering rates decrease significantly. At the
                                                                   bottom of the Anaconda Ponds (depth  26-28 m), the predicted
                                                                   metal concentrations are less than at the bottom of the Opportunity
                                                                   Ponds (12-14 m). This is because  the leaching period is approxi-
                                                                   mately twice as long at the bottom of the Anaconda Ponds, and
                                                                   approximately the same quantity of metal is being leached. Solute
                                                                   concentrations predicted to leach from various zones of the tailings
                                                                   ponds are shown in  Table 3.
                                                                                              Table 3
                                                                             Predicted Metal Concentrations and Durations
                                                                         Of Release from Various Zones Within the Opportunity
                                                                         And Anaconda Ponds Interaction of Acidity and Metals
                                                                               With Unsaturated and Saturated Alluvium

Zone Av<
1
i
3
4
5
6
7
8
9
10

erige Thickness
1 Tillfngs, m
12
31
5
4
9
7
9
a
7
B

Oxidizing Zone Retches
Bottom, yrs
>60.000
> 60. 000
4,000-6,000
3, 000- S, 000
10,000-12,000
8,000-10,000
10,000-12,000
9,000-10,000
8,000-10,000
9,000-10,000


0.12-0.26
0.12-0.26
0.06-0.10
0.04-0.10
0.08-0. IS
0.08-0.12
0.08-0. IS
0.08-0.12
O.OB-0.12
0.08-0. 12


0.1-1.0
0.1-1.0
O.i-1.0
0.1-0.9
0.3-1.0
0.3-1.0
0.3-1.0
0.3-1.0
0.3-1.0
0.3-1.0


2S-3SO
25-3SO
3S-S10
30-420
4S-6SO
40-600
45-650
40-600
40-600
40-600


0.06-0.10
0.06-0.10
0.06-0.12
0.06-0.12
0.07-0.12
0.08-0.12
0.08-0.12
0.08-0.12
0.08-0.12
0.08-0.13
'I.......

10-220
10-220
14-300
12-240
18-380
16-320
18-380
16-320
16-320
16-320
 Arsenic
 Cadmium
 Copper
 Lead
 Zinc
                  0.03-0.05
                   0.2-0.9
                    25-380
                  0.04-0.08
                    10-220
                                           0.02-0.05
                                            0.7-1.6
                                             51-339
                                           0.06-0.1
                                             49-201
  The geochemical equilibrium model MINTEQ was applied at
selected pH values in the oxidizing zone to determine whether the
predicted  aqueous  phase concentrations are  near or exceed
saturation.
  Based on MINTEQ predictions, the concentrations predicted in
Table 2 are below saturation. Hence, the movement of aqueous
phase metals out of  the zone of active oxidation appears  to be
limited by the rate at which they are weathered into the aqueous
phase (which depends on pH), the infiltration velocity through the
oxidized zone and adsorption characteristics of the solutes within
the zone rather than on saturation limitations.
  As  metals are transported out of the oxidizing zone,  they
reprecipitate as secondary minerals and adsorb on the underlying
tailings. However, as long as the pH in the oxidizing zone is low
enough so that solubility products are not exceeded, redissolution
will occur as the active zone moves downward.
  The mass of aqueous phase metals M(z) passing through any
horizontal plane in the tailings  can be  approximated by:
INTERACTION OF ACIDITY AND METALS
WITH UNSATURATED AND SATURATED ALLUVIUM
  Silicate minerals in the tailings ponds neutralize most of the
acidity before it leaves the ponds. Consequently, the effective depth
of the tailings ponds, as far as acidity release is concerned, is ap-
proximately
2 m when the active oxidized zone is in contact with the under-
lying alluvium.
  Next, the downgradient distance, x, required to neutralize the
M(z) =  , E , C*i u,
                                                        (10)
                                                                                              Figure 8
                                                                              Idealized Cross-Section of the Tailings Ponds
                                                                             Used to Predict Distance x to Consume Acidity
                                                                                      Generated from the Pond.
                                                                                                     MINING WASTES     447

-------
acidity that does leave the tailings ponds can be calculated (Fig. 8).
This calculation can be done by developing an equation based on
the following relationship:

   Acidity remaining in tailings = carbonate alkalinity in unsaturated
                             alluvium below tailings

                             carbonate  alkalinity in saturated
                             alluvium below and down-
                             gradient of tailings

                             alkalinity in  flowing
                             groundwater

   Using typical data from the site, if the unsaturated zone beneath
 the tailings is 2.5 m or greater, then the  acidity released  by the
 tailings is completely neutralized within this zone. For those loca-
 tions where the tailings are in direct contact with the groundwater,
 the acidity is predicted to be consumed within approximately 0 to
 2400 m downgradient, depending on how quickly the leachate mixes
 with the groundwater.
   As metals are transported into the unsaturated zone beneath the
 tailings ponds, precipitation initially occurs. If the zone is thick
 enough (i.e., greater than approximately 2 m), the acidity in the
 tailings is consumed, and the metals can be precipitated in the
 alluvium. In the zone of deposition in the unsaturated alluvium,
 the metal solubilities predicted by MINTEQ are shown in Table 4:
                                                         Consequently, the unsaturated and near surface saturated zones
                                                      beneath the tailings ponds are predicted to become enrichment zones
                                                      and to greatly decrease the aqueous phase concentrations of cad-
                                                      mium, copper and  zinc as they enter the groundwater.
                                                      CONCLUSIONS
                                                        Increased concentrations of arsenic, cadmium, copper, lead and
                                                      zinc are predicted to be generated within the tailings ponds at
                                                      Anaconda. Over a period of thousands of years, these solutes are
                                                      predicted to slowly leach deeper into the tailings, while aqusous
                                                      phase concentrations are gradually enhanced.
                                                        Due to the considerable amount of alkalinity contained in the
                                                      underlying alluvium, the capacity exists to neutralize the acidity
                                                      released from the tailings and to precipitate most of the metals in
                                                      an unsaturated zone above the groundwater if the depth of the un-
                                                      saturated zone is 2.5 m or thicker.  Residual acidity that does the
                                                      reach the groundwater is predicted to be rapidly consumed, further
                                                      attenuating solute concentrations. Consequently, while elevated
                                                      metal concentrations are being generated within the tailings, their
                                                      impact on the groundwater will be highly localized.
 Metal
                           Table 4
             Metal Solubilities Predicted by MINTEQ
 Arsenic
 Cadmium
 Copper
 Lead
 Zinc
Concentration
   Entering
Alluvium (mg)

     0.2
     1.0
   300-600
  0.08-0.12
    20-320
  Concentration
      after
Precipitation (mg)

      0.2
      0.01
      0.6
   0.08-0.12
      0.5
Controlling
     Solid

none
CdCOj
Malachite
none
ZnSiO3
REFERENCES

1. Gherini, S.A., Mole. L., Hudson RJ.M., Davis, G.R, Chen, C.W. and
  Goldstein, R.A., "The ILWAS Model- Formulation and Application;'
   Water. Air. and Soil PolluL 26,  1985, 425-459.
2. Lapidus, L. and Finder, G. F. Numerical Solution of Partial Differen-
  tial Equations in Science and Engineering. John Wiley & Sons, New
  York, NY  1982.
3. Singer, P.C. and Stumm,  W. "Kinetics of the  Oxidation of Ferrous
  Iron." In:  Proc of the 2nd Symposium on Coal Mine Drainage.
  Bituminous Coal Research, Monroeville, PA., 1968, 12-34.
448    MINING WASTES

-------
                  Technical  Approaches  on  the UMTRA  Project
                                Relevant to  Superfund  Projects
                                                  Jack A. Caldwell
                                                    Duane Truitt
                                         Jacobs Engineering Group  Inc.
                                            Albuquerque, New Mexico
ABSTRACT
  The Uranium Mill Tailings Remedial Action (UMTRA) Project
is a major Federal undertaking to reclaim 24 inactive uranium mill
tailings piles in 10 states. In the 5 years that work has been in
progress, numerous  technical and  programmatic standard
approaches have been developed and several advances have been
made. This paper discusses the UMTRA approaches, the parallels
between UMTRA and Superfund and the UMTRA advances that
are relevant to Superfund remedial activities. Included are descrip-
tions of an infiltration model to simulate unsaturated flow through
covers; methods to select rock for erosion barriers which must be
stable for long periods and resist disturbance by large floods; and
descriptions of  programmatic controls  and management
approaches.

INTRODUCTION
  The U.S. Department of Energy (DOE) Uranium Mill Tailings
Remedial Action (UMTRA) Project is performing closure and
reclamation at 24 inactive uranium mill tailings piles in 10 states.
Remedial action includes site  characterization,  environmental
assessments,  preparation  of remedial action  plans, construction
and long-term surveillance and maintenance.
  While the laws, regulations and standards  that  govern the
UMTRA and Superfund Projects differ, there are, nevertheless,
significant areas of correspondence between the two with regard
to ultimate objectives, technologies and appropriate technical
approaches. In completing work at 4 sites and planning cleanup
at the remainder, the DOE and its contractors  have pioneered
numerous technical innovations. This  paper discusses  parallels
between the programs and technical approaches and advances on
the UMTRA Project which may be relevant to Superfund sites.

PROJECT DESCRIPTIONS: UMTRA
  Technical work on the UMTRA Project is performed according
to the Technical Approach Document (TAD)1. The TAD details
acceptable design approaches and methods of analysis in the areas
of:

• Geotechnical, hydraulic and  hydrologic engineering
• Groundwater assessments
• Radiological Health and Safety

  The NRC's Standard Review Plan  (SRP) describes the data,
design methods and checklists of details sought in their review of
Remedial Action Plans (RAPs).
  UMTRA Project site characterization seeks full definition of
regional and site geology,  surface water, groundwater contamina-
tion and other environmental (including socio-political) aspects of
the site. Site characterization reports and National Environmen-
tal Policy Act (NU.S. EPA) documents are prepared.
  Remedial action at UMTRA Project sites usually consists of
either stabilization in place or relocation of the tailings to a better
site. If the tailings are relocated, liners or geochemical barriers may
be placed first at the new disposal site, followed by the tailings.
If the tailings are not relocated, the pile is shaped to promote run-
off. In either case, the pile is covered with a low-permeability barrier
(to reduce radon emanation and groundwater infiltration) and
erosion resistant rock and provided with surface drainage features.
The design is described in the RAP. Figure 1 shows a typical
UMTRA pile remedial plan.
   BEFORE RECLAMATION
   Erosion Barrlar
   Radon
   Barrlsr
Wndblown Contaminated Soils

       Windblown Soil!
   AFTER RECLAMATION (Mellon A-A'I
       1'-? Erosion Barrltr
        • Durabla Rock
       6' Bidding  • Clssn Sand
       X Radon Barrlar
        • Compacted Clay
         4 Silt
       Tailings
        • Sand, SIM. ป Clay
   COVER DETAIL
                               Original Parlnwtsr of Plla - Tailings
                               Ralocatsd to Fill Pond Araa -
                            PLAN LAYOUT
                          Figure 1
                 Typical UMTRA Pile Layout
                                                                                                MINING WASTES    449

-------
   After construction, surveillance and maintenance programs are
 established2. This  program  usually involves  semi-annual  and
 annual inspections and the completion of required corrective action.

 PROJECT DESCRIPTION: SUPERFUND
   CERCLA established a remedial program for inactive hazardous
 waste sites. The NCP. contained  in 40 CFR  300, outlines the
 requirements of the CERCLA remedial program. The NPL was
 established to identify those sites considered to pose the greatest
 and most immediate threat to public health and the environment.
   The original Superfund law authorized the program to operate
 for  5 years. When that authorization expired in 1985, Congress
 provided interim funding until SARA of 1986 was  enacted in
 October 1986. Besides providing an additional 5 years of authori-
 zation, SARA was intended  to correct several  flaws or omissions
 in the original act. The 1980 act had not specified any standards
 for  cleanup of sites; the issue of "how clean is clean?" had not
 been addressed. Congress adopted the use of Applicable or Relevant
 and Appropriate Requirements (ARARs) which the U.S. EPA must
 identify and implement for remedies under SARA. SARA also re-
 quires the President to select, to the maximum extent practicable,
 remedies that use permanent solutions and alternative treatment
 technologies or resource recovery technologies. Furthermore, the
 selected remedies should reduce permanently the mobility,  volume
 and toxicity of contamination at Superfund sites. Excavation of
 wastes and transport to another landfill without treatment is ex-
 pressly given the least preference  when other  treatment alterna-
 tives are practicable.

 COMPARISON OF THE UMTRA AND
 SUPERFUND PROJECTS
   Both the UMTRA Project and Superfund are major Federal sta-
 tutes involving remedial action at sites containing materials poten-
 tially hazardous to human health. Some differences between the
 programs include:

 •  UMTRA  Project sites contain  large volumes (up to  several
   million cubic yards each) of materials with relatively low con-
   centrations of contamination (typical constituents are radon gas
   precursors, acids, caustics  and sulfates), while Superfund sites
   typically contain  smaller volumes of contamination (such as
   heavy metals, organic solvents, PCBs, etc.),  sometimes  at very
   high concentrations.
 •  Uranium mill tailings are fairly consistent from site to site, thus
   DOE remedial actions share many similarities. Superfund site
   wastes vary greatly, hence most remedies are very site, specific.

   The most important parallel between the  two programs is the
 requirement for  long-term  (i.e.,  "permanent") remedies.  On
 UMTRA, projects,  the requirement to  provide for a  design life
 of 1,000 years where practical and nevertheless for  200 years,
 represents a societal choice of a de facto permanent remedy. Other
 similarities in approach justify the attention of persons involved
 in Superfund work. For example:
 •  Some of the NPL sites contain mine and mill wastes including
   uranium mill tailings.
 •  Land disposal without treatment  is almost banned under SARA;
   nevertheless, wastes treated to reduce permanently the mobility,
   volume or toxicity of contaminants still may be disposed in a
   landfill. Landfill  wastes in which the contamination  risk is
 reduced, but not eliminated, must be stable against dispersal by
   wind or water. Therefore, the need to model contaminant trans-
   port (particularly via groundwater) would continue. Thus, the
   technical approaches developed on the UMTRA Project may be
   directly applicable to post-SARA remedies.
 •  Many technical and administrative procedures developed for the
   UMTRA Project (ie., value engineering reviews, quality assur-
   ance planning, quality control of technical data, project integra-
   tion and controls, etc.) may be adapted to  Superfund  use.
   The following sections of this paper discuss technical approaches
used on the UMTRA Project that may be applicable to the Super-
fund  remedial program.

INFILTRATION
   A common approach to analyzing infiltration and run-off from
a facility over which a cover is placed (or proposed) is the HELP
computer code. On the UMTRA Project, a considerably  more
sophisticated code called SOILMOIST has been developed and is
used.
   The SOILMOIST code was developed in response to criticism
that analysis of infiltration for saturated conditions is unrealisti-
cally conservative for application under unsaturated conditions.
Where the cover consists of a thick radon/infiltration barrier of
compacted clay, a bedding or filter layer of permeable sand and
the erosion resistant rock layer, two parameters are of significance:
the long-term moisture content of the radon barrier (radon flux
through the barrier is very sensitive to soil moisture content) and
the infiltration through the barrier to the tailings and hence to the
groundwater.
   The SOILMOIST code starts with a file of daily precipitation,
temperature and evaporation  at the site. These data are derived
from  records or simulated by climatic generation subroutines.
Normally the model is run to simulate conditions over a relatively
long period, 30 to 100 years. Time steps are usually set at 1  day.
The model determines climatic conditions for each day.  If it rains
that day, the model calculates the volume of precipitation that flows
downslope over the relatively impermeable radon barrier (a
function of the pile slope and the hydraulic conductivity of the
radon barrier and bedding layers). Also calculated is the volume
of precipitation that enters the radon barrier. Flow through the
radon barrier is modeled for both saturated and unsaturated con-
ditions by specifying, for the radon barrier soil, the relationship
between soil moisture content and hydraulic conductivity. If rain
continues the next day, infiltration will continue through the radon
barrier. If the next day is dry, evaporation from the barrier soil
and through the bedding and rock will occur and an upward capil-
lary gradient may be established. The model simulates  and  ana-
lyzes such conditions. Figure 2 shows the essential components of
                                      • now WTENMTV-OUIATION
                                            noon.
                                      • HATE OF FUJW TMWUOM FUC*
                                      • EWOFIATION TMFJOUOM PUBt
                                       t NOCK

                                      MAOOH MMMII
          .       .      ..
           *• .   * 7 -\ * • *  7 g '•'•
          '
 • SEEPAOERATE
 • HVDRAUUC cowucnvTTY
 • OEONEC OF SATURATION

 WATER CONTAMHATKM
 • TARJHOS CONSmUENTป

I VAOOM ZOM HOOn.
 • SEEPAOE RATE
 • ATTENUATO*
           •>.•.'.
                    O '  '
                   	ป  . -
 • TRANSVERSE FUJW RATE
 • in ma
     LTAMT WATER QUALITY

   WD             •
   I HOCK   r^ป oun
                          Figure 2
SOILMOIST Model and Groundwater Impact Assessment Methodology
450    MINING WASTES

-------
the model and the manner in which they integrate with a compre-
hensive assessment of the impact of infiltration and seepage on
groundwater quality.
  For example, the authors have proposed using the SOILMOIST
model at a Superfund tailings site because of the arid climate of
the site and the need to analyze partially saturated infiltration to
accurately define the quantity of seepage through  the tailings to
the groundwater. In addition, the model will be able to define the
long-term (i.e., 1,000 years) hydrologic conditions at the site.


ROCK EROSION PROTECTION
  U.S. EPA standards require that the tailings be stabilized against
dispersal by wind or water for 1000 years to the extent reasonably
achievable and, for at least 200 years. The NRC3  has stipulated
that the design hydrologic events for uranium tailings shall be the
Probable Maximum  Precipitation (PMP)  and  the  Probable
Maximum Flood (PMF).
  The TAD1 outlines procedures to design erosion protection for
tailings piles. Prediction of flood parameters, such as maximum
volumetric flow rate, velocity, direction and depth of  flow, is
accomplished via commonly used hydrologic and hydraulic analysis
tools such as the HEC-1 and HEC-2 computer codes. The hydraulic
parameters are input to the  rock sizing calculation.
  The PMP incident on the pile is adjusted for the relatively small
watershed areas of tailings piles (with resultant rainfall intensities
approaching 50 to  100 in./hr at some sites) and is routed as sheet
flow across the pile top and sideslopes. The resulting hydraulic
parameters are velocity and depth of flow, which are input to the
rock sizing calculation.
  The rock sizing calculation used on the UMTRA Project is the
Safety Factors Method4.  This analytical procedure considers the
shear stress exerted upon a particle by flowing water. Any shear
stress greater than the critical shear stress would cause a particle
to become dislodged  by  the flow. The  critical  shear stress is a
function of both hydraulic forces and the characteristics of the
particle such as size, density, shape and resistance to flow. The
name "Safety Factors" refers to a calculational approach whereby
a minimum safety factor is input and the minimum required rock
size is output. For the UMTRA Project, the factor of safety for
PMP/PMF events is 1.0; thus, for any lesser event, the  factor of
safety is  greater than 1.0. Since a particle's resistance to flow is
a function of size and shape and the resistance  to flow of a bed
of similar particles  affects hydraulic parameters such as flow depth
and velocity (which in turn affects the hydrologic time of concen-
tration and  storm intensity),  this method  requires an  iterative
approach to solution.
   A limitation of the Safety Factors method is that it is accurate
only on slopes of 10% or less. That limit is acceptable for tailings
pile top slopes,  which are usually specified as the minimum neces-
sary to provide positive drainage (usually 2 to 4%); however, space
limitations and the desire to limit the surface area usually dictate
the use of much steeper pile sideslopes (typically 20% or 5 to one).
Stephenson5  proposed an empirical method based on  flume
studies by Oliver6 for use  on  slopes of 2 to 20%.  That method has
been adopted for UMTRA designs.
  A typical  rock  erosion protection design  resulting  from the
UMTRA Project procedures might consist of 1 ft of small, diameter
rock (median diameter of 1.5 to 2.0 in.) on the top slopes of the
pile, with 1.5 ft of larger rock  (median diameter of 6 to 8 in.) on
the steeper side slopes. Both layers are  underlain with granular
filters as necessary to prevent piping of the much  smaller under-
lying soil particles. Channels and berms usually are not used in
the design of piles  in order to avoid the resulting flow concentra-
tions. If a channel must be used, it is armored with riprap designed
using the methods given. Of particular concern, especially in the
arid western states (which have sparse vegetation), is gully intru-
sion. At  sites where geomorphological analysis indicates a poten-
tial for gully intrusion, the heads of existing gullies may be armored
with riprap, or the toe of the pile may be protected with an apron
of riprap.
  For example, one Superfund site with which the authors are
associated is in the floodplain of a major river in the midwest. The
proposed remedy includes stabilization of a landfill against flood
damage. The conceptual design features  riprap which could be
designed using the UMTRA methods.

ROCK DURABILITY
  The durability of rock erosion protection is a significant design
factor because of the 1000-year life of UMTRA Project tailings
piles. Rock durability (defined as the ability to withstand the forces
of weathering) affects the long-term ability of rock particles to with-
stand erosive  forces. The TAD1 method of considering rock dura-
bility in remedial designs is to characterize the available rock at
nearby borrow sites and, using durability test data, oversize the
rock to maintain  the required erosion protection for 1000 years.
The specified rock durability tests include:

  Bulk Specific Gravity and Absorption  (ASTM C127)
  Petrographic Examination (ASTM C295)
  Sulfate  Soundness (ASTM C88)
  Hardness (point load or Schmidt Rebound Hammer)
  Los Angeles Abrasion (ASTM C131  or C535)

  Composite scores resulting from the above tests are assigned to
given sources of rock3. The  least  costly source of "good" rock
(scoring at least 80 out of a  possible 100  points) is located. The
rock is oversized  according to the following criteria7:

                 Frequently Saturated Areas
                  (Channels, Toes, Aprons)
           Score
Oversize Factor
           80- 100
           65 -  80
    2
    10
    Reject
                Occasionally Saturated Areas
                (Top Slopes and Side Slopes)
           Score
Oversize Factor
           80- 100
           65 -  80
    no oversizing
    2
    Reject
   A 1,000-year design life might not be a requirement under the
 Superfund Program, but the need to provide a permanent remedy
 would certainly make the design method described above a prudent
 design course.
 PROGRAMMATIC
   The UMTRA Project is  a large multi-site undertaking. Site
 characterization has been performed at each of the 24 sites. In
 addition, NU.S. EPA documentation and Remedial Action Plans
 (discussed previously) will be completed at each site. Other program
 activities include:
 • Compilation of construction drawings and specifications
 • Preparation and implementation of Health and Safety Plans,
   QA/QC programs, construction monitoring and audit programs
 • Post-construction plan compilation and surveillance and main-
   tenance  programs

   All of these activities have direct parallels  in the Superfund
 program as outlined in the NCP. Such Superfund activities would
 include preparation of remedial investigations/feasibility studies,
 remedial design and construction. In completing the above activi-
 ties on the UMTRA Project, numerous documents have been com-
 piled. The references and bibliography sections at the end of this
 .paper list  the  pertinent guidance documents  for the UMTRA

                                  MINING  WASTES    451

-------
Project as well as other relevant references.
  Management  of the UMTRA  Project involves coordinating
numerous Contractors such as the Technical Assistance Contrac-
tor (TAC) and the Remedial Action Contractor (RAC). In addition,
it is necessary for management to track progress on the many varied
phases of the 24 sites. In order to plan, track and control the
Project,  an Integrated Project Management and Control System
has been established and is operated by the TAC for the DOE.
The basic system  is a  variant of the Cost  Schedule Control
System8. While Superfund projects may differ in their size, com-
plexity and management and control needs, the approaches adopted
on the UMTRA Project may be adapted to the needs of many
Superfund activities.


CONCLUSION
  This paper  has set out salient  aspects of one major Federal
program—UMTRA—that  are  relevant  to  another Federal
program—Superfund. Differences between the 2 programs should
not mask similarities and the value of technical methods which are
useful on one program that could  be applied to the other. Cross-
fertilization of  ideas,  approaches,  methods  and philosophies
advances societies and their technologies; this paper has attempted
to contribute to that cross-fertilization by describing infiltration
analyses, rock durability and sizing approaches and programmat-
ic methods used on the UMTRA Project. It remains for engineers
now to advance the use of these approaches in relevant situations
on Superfund.


REFERENCES
 1.  U.S.  Department of  Energy, "Technical Approach  Document,"
    UMTRA Project Office, Albuquerque Operations Office, Albuquer-
    que, NM, May 1986.
 2.  U.S. Department of Energy, "Guidance for UMTRA Project Surveil
    lance and Maintenance," UMTRA Project Office, Albuquerque Oper-
    ations Office, Albuquerque,  NM,  January 1986.
 3.  U.S. Nuclear Regulatory Commission, Methodologies for Evaluating
   Methodologies for Evaluating Long-Term Stabilization  Designs of
    Uranium Mill Tailings Impoundments, Division of Waste Management,
   Office of Nuclear Material Safety and Safeguards, U.S. Nuclear Regula-
   tory Commission, Washington, D.C., NUREG/CR-4620, June, 1986.
 4. Stevens, M.A., D.B. Simons and  G.L. Lewis, "Safety Factors for
  Riprap Protection," J. Hyd. Eng.,  1976
 5. Stephcnson, David, Rock fill in Hydrologic Engineering, Elsevier Scien-
   tific Publishing Company, New York, NY,  1979.
 6. Olivier, "Through and Overflow Rockfill Dams New Design Tech-
   niques," in Proceedings of ICS, published in Rockfill Hydraulic En-
   gineering, Elsvier Publishing, New York, NY, 1967.
 7. U.S. Nuclear Regulatory Commission, draft letter from T. Johnson
   concerning design procedure for rock January 29, 1987.
 8. DOE "Cost and Schedule Control Systems Criteria for Contract Per-
   formance Measurement," Office of Project Facilities Management,
   U.S. Department of Energy, Washington DC, Apr. 1984
BIBLIOGRAPHY
Agogino, K., R. Portillo, and B. Keshian Jr., "Design of Drainage Facili-
     ties for the UMTRA Project," Eighth Annual Symposium of Geotech-
     nical and Geohydrological Aspects  of Waste Management,  Fort
     Collins, CO, Feb. 1986.
Brinkman. J. E.. W. A. Ericson, J. B. Price and D. Lewis, "Hydrologic
     Setting and Groundwater Chemistry Characterization of the Vitro
     Uranium Mill Tailings Site South Salt Lake, Utah," Sixth Symposium
     on Uranium Mill Tailings Management, Fort Collins, CO, Feb. 1984.
Caldwell, J. A. and A. M. G. Robertson, "An Evaluation of Geotech-
     nical and Hydrologic Aspects of Uranium Mill Tailings Reclamation
     in the USA  and Canada,"  Presented at Waste Management '87
     Tucson, AZ, Mar.  1987.
CaldweU, J. A. and T R. Wathen, "Criteria for Remedial Works at Inactive
     Uranium Mill Tailings Piles," Second International Conference on
     Radioactive Waste Management, Winnipeg, Manitoba, Canada, Sep.
     1986.
D'Amonio, J. R., J. A. Caldwell and G.R. Thiers, "Burying the Nuclear
     Past," Civil  Eng.. Feb. 1987.
Deutsch, W. J. and J. W. Thackston, "Aquifer Restoration Considera-
     tions at Inactive Uranium Mill Tailings Sites," Proceedings of the
     Fourth  Annual  Hazardous Materials  Management Conference,
     Atlantic City. NJ. June 1986.
Dupuy, J. R., W. J. Deutsch, J. Hilton, G. Rice and J. W. Thackston,
     "Water Resource Protection and the UMTRA Project: Three Case
     Histories,"  National Waterwell Association  Conference in South-
     western Ground Water Issues.  Tempe, AZ, Oct.  1986.
Knight, W. C., V. Fry and R. Sena,  "Conceptual Design of Remedial
     Actions at Three Na*. ajo Sites," Second International Conference on
     Radioactive Waste Management, Winnipeg, Manitoba, Canada,  Sep.
     1986.
Larson, N. B. and B. Mitchell, "Cone Penetrometer Use on Uranium Mill
     Tailings," In-Situ '86, a specialty conference sponsored by theGeo-
     technical Engineering Division of the American Society of Civil
     Engineers, Blacksburg, \.\,  June 1986.
Mason. W. C. and R. L. Williams, "Challenge of Canonsburg-A Case
     History," Waste Management 198) Conference, Tucson, AZ, March
     1985.
Nelson, R. A., W. J. Smith and K. R. Baker, "The Range and Variability
     of Radium Concentration and Emanating Fraction in  Uranium Mill
     Tailings and Their Impact on Radon Barrier Design," Waste Manage-
     ment 1985 Conference, Tucson, AZ, March  1985.
Peterson, S. R. and W. J. Deutsch, "GcochemicaJ Reaction Modeling in
     Contaminant Migration Studies of the Uranium Mill Tailings Environ-
     ment," National Waterwell Association Conference  on Solving
     Groundwater Problems  with Models, Denver, CO, Feb. 1987.

Rager, R., J. A. CaldweU and D. Harvey, "Seismic Design Criteria for
     the UMTRA Program: Their Influence on Pile Stabilization Proce-
     dures," Abstract of a paper submitted to the Fifth Canadian Con-
     ference on Earthquake Engineering,  Ottawa,  Ontario July 1987.
Smith, W. J., R.  A. Nelson  and K. R. Baker, "Sensitivity Analysis of
     Parameters Affecting Radon Barrier Cover Thiclcness," Seventh Sym-
     posium on Uranium Mill Tailings Management, Fort Collins,  CO,
     Feb.  1985.
452     MINING WASTES

-------
                         Utilization  of  Mobile Incineration  at  the
                           Beardstown  Lauder  Salvage  Yard Site

                                              James A. Janssen, P.E.
                                                   Robert Munger
                                       Division of Land Pollution Control
                                    Illinois Environmental Protection Agency
                                                 Springfield, Illinois
                                               John W. Noland,  P.E.
                                                 Nancy P.  McDevitt
                                              Luis A. Velazquez. P.E.
                                                Roy F. Weston, Inc.
                                            West Chester, Pennsylvania
ABSTRACT
  In October 1986, the Illinois EPA (IEPA) contracted with Roy
F. Weston, Inc. (WESTON) to construct and operate a mobile
incineration system at the Beardstown Lauder Salvage Yard in Cass
County, Illinois. This mobile incinerator,  when built will treat
between 5,000 and 10,000 tons of soil contaminated with PCBs.
  A description of the site history, past cleanup and sampling
activities, site contamination characterization, permit requirements,
incineration process description, emissions monitoring procedures
and incinerator operating plans and schedule will be discussed. Par-
ticular attention will be given to the U.S. EPA Toxic Substances
Control Act (TSCA) and IEPA permit requirements with emphasis
on the implementation of the trial burn plan.
  Mobilization of the 5 ton/hr rotary kiln incinerator is complete
with shakedown occurring in August 1987. The trial burn will take
place in mid-September 1987. After the trial burn, the incinerator
will be shut down pending laboratory analysis and U.S. EPA and
IEPA review and approval of the trial burn  data.
  After determining that the incinerator has  met stringent state
and Federal performance standards and that  operation of all equip-
ment has been deemed satisfactory,  full-scale operation will  be
permitted. If all goes well, operations will commence in December
1987.

INTRODUCTION
  Until 1974, Russell Lauder operated a salvage yard near Beards-
town, Illinois. This business involved the stockpiling of metal scrap,
including vehicles, white goods, electrical equipment and other used
metal goods. In addition, large stockpiles of  tires were kept on-
site. From 1972 to 1974, the environmental agencies received several
complaints from salvage yard neighbors of  open burning that
created a smoke and odor nuisance. One the purpose of the fires
was  to increase the scrap value of discarded copper wire  by
removing insulation. The electrical equipment taken to the salvage
yard included transformers and capacitors. The copper coils were
removed for resale while the remaining outer  casings and fluids,
some containing PCBs, were allowed to accumulate with no attempt
to identify PCB materials or to take precautions to prevent leakage.
  After Mr. Lauder's death in  1974, the  Illinois EPA (IEPA)
dropped pending legal action to stop the  open burning.  IEPA
records show no further complaints of open  burning.
  In April 1984, a former salvage yard employee informed the
IEPA that PCB-contaminated materials had been present on-site
and that some had been dumped onto the ground during copper
reclaiming activities. To verify that PCBs were actually present,
nine soil samples were taken during two site visits in April and July
 1984. The highest concentration was 120,000 ppm (12%), which
 was found in a sample taken from soil saturated with oily waste.
  After confirming that PCBs were present, the IEPA attempted
 to identify those responsible for the PCB waste. The investigation
 found that a local bank had accepted an interest in part of the land
 to settle a debt owed by the site owners. These parties were unable
 or unwilling to pay for a detailed study of the contamination and
 removal of PCB articles and contaminated soil. The generators
 of the waste also could be held responsible for cleanup costs if they
 could be identified. Therefore, the IEPA attempted to discover
 the sources of the PCB electrical equipment. All legible serial
 numbers on transformers and capacitors were recorded, with the
 hope that the manufacturers of the equipment would have records
 of who bought  the equipment. Unfortunately, in every case, those
 records  had been discarded.
  With the current owners of the property unwilling or unable to
pay cleanup costs and the identity of the waste generators buried
in the past, the IEPA realized that public funds would have to be
used  for remedial action. Initial  surveys revealed that detailed
sampling could not proceed without first removing the large amount
of metal scrap  and other debris on  the site. Given the need to
commit  state funds to the project and the need to remove large
amounts of non-hazardous debris before characterization of the
soil contamination could be done, the Agency devised a four-phase
program to manage the site. On May 1, 1985, IEPA Director
Richard Carlson signed a Record of Decision outlining the sequence
of remedial action events that follows:

• Phase I  - Removal of all nonhazardous debris, transformers,
            capacitors and visibly  contaminated soil
• Phase II - Detailed soil sampling to determine the amount of
            soil to be removed
• Phase III- Removal of contaminated soil identified in Phase
            II to a licensed PCB landfill
• Phase IV- Groundwater  monitoring

  On May 16,1985,17 cleanup contractors attended a pre-bid con-
ference at the site. On May 23, the IEPA issued a formal Request
for Quotation to these contractors, setting forth bidding require-
ments. By June  6, the IEPA had received four proposals addressing
all requirements in the Request for  Quotation.
  Mid-America Environment Service, Inc. submitted the lowest
bid of the four technically adequate proposals. Mid-America began
on-site activities on June 24.  By July 25, removal of the debris and
visibly contaminated soil was complete. Four thousand yards of
uncontaminated scrap metal had been taken to IBS, Inc., a Peoria
metal scrap dealer. Twenty overpack drums containing PCB trans-
                                                                                                INCINERATION    453

-------
formers, capacitors and liquids were taken to a PCB incinerator
operated by Rollins Environmental Services in Deer Park, Texas.
One hundred twenty yards of contaminated soil were shipped to
the Chemical Waste Management, Inc., landfill in Alabama. One
hundred twenty five yd3 of tires were taken to the Sangamon Valley
landfill near Springfield, Illinois. Later in 1985, 3,000 gal of non-
hazardous washwater from decontamination of personnel  and
equipment were taken to Chem-Clear, a Chicago facility that treats
aqueous waste.
  Cleared of scrap metal and other debris,  the site was accessible
to the equipment needed for detailed representative sampling. Phase
II of the remedial action was a detailed characterization of the PCB
contamination remaining after Phase I. Agency employees sampled
soils using a hollow-stem auger drill rig and a split-spoon sampler.
In 54 locations, composite samples from to 2 ft and 2 to 5 ft were
taken. In five of these locations, additional samples composited
from material taken from 5 to 7.5 ft and 7.5  to 10 ft were taken.
No PCBs were found in the samples  taken from  depths greater
than 5 ft. At 55 points, samples consisted of material taken from
a depth of 0 to I ft. PCBs were detected in 23 samples at 22 different
locations. The analytical work done showed  that on most of the
site PCB concentrations were below the detection limit of 5 ppm.
Where PCBs could be detected, concentrations ranged  from 5 to
1,690 ppm.
  The contamination is believed to be limited to the top 5 ft of
soil; no PCBs have been detected in groundwater samples taken
from monitoring wells around the site and private water supply
wells. The Phase II characterization, however, showed that PCBs
remain in surface soils at unacceptable concentrations.
  The original ROD signed by Director Carlson called for the con-
taminated soils identified in Phase II to be removed to  a licensed
PCB landfill. However, destruction of the PCBs was preferred to
simply moving them to  a site which  someday also may require
remedial action. Unfortunately, existing incinerators are operating
at full capacity. Therefore, the IEPA  revised the remedial action
strategy in the original ROD, substituting thermal destruction in
a mobile incinerator for landfill disposal. This change had the
additional advantage of eliminating the shipping of contaminated
soil on public highways.
  The IEPA solicited technical and cost proposals from qualified
contractors to construct and operate the mobile incinerator at the
Lauder Salvage Yard site. On October 4, 1986, the IEPA awarded
the contract to Roy F. Weston, Inc.

PERMIT REQUIREMENTS
  Three permits are required for operations at the Lauder Salvage
Yard:

• U.S. EPA National TSCA permit
• IEPA Division of Land Pollution Control permit to develop a
  treatment facility
• IEPA Division of Air Pollution Control permit to construct and
  permit to operate

  The Transportable Incineration System  (TIS) will be the first
full-scale TIS in the country to receive a national TSCA permit
valid in any U.S. EPA region. Acquisition of this permit will avoid
the redundancy associated with repermitting the unit as it moves
throughout the State of Illinois or from region to region.

PROCESS DESCRIPTION
  A schematic of the process flow diagram is shown in Figure 1.
The TIS has the capability to process solid wastes. Classified soil
(i.e., less than 2 in. in diameter) is introduced into the solid  waste
feed  system by a front-end  loader. The solid waste feed system
consists of the following components:

  • Dump hopper/withdrawal screws
  • Cross-drag conveyor
  • Rotary kiln feed screws
  The solid waste feed system is designed to accommodate 12,000
Ib/hr of soil (wet  basis).
  The rotary kiln is a rotating cylindrical steel shell which is refrac-
tory, lined and mounted at a slight  incline  from the horizontal
plane.  The rotation provides  movement  of the feed material
through the kiln mixing of material with combustion air. A burner
is mounted above the feed inlet. The burner flame is concurrent
with the movement of the feed material. A photograph of the rotary
kiln taken during  fabrication is shown  in  Figure 2.
  The rotary kiln is designed to provide a minimum soil retention
time of 30 min. The kiln is  designed to  dry and to heat the feed
material to 1,800ฐF; however, lower  operating temperatures may
be required to avoid material slagging.  The actual  operating
temperature will be selected to ensure that the ash contains less
than 2 ppm per resolvable  chromatographic peak in  the  PCB-
retention  window  and  5  ppm total PCBs (actual  operating
temperature will be  established during the  demonstration trial
burn). Natural gas or propane will be used  to fuel the incinerator;
however, future modifications may include conversion to fuel oil.
  The heated ash  discharging the rotary kiln is cooled to 300 T
with conditioning  water sprays and deposited  into watertight
dumpsters (approximately 12-yds capacity).
  Exhaust gases from the rotary kiln pass through a secondary
combustion chamber (afterburner). This chamber is a stationary,
vertically-mounted,  cylindrical, refractory-lined steel shell.  Its
burner is horizontally-mounted, perpendicular to the path of the
flue gas. The chamber is designed to provide a minimum gas  reten-
tion  time of 2 sec. Natural  gas or propane is used to heat gases
in the secondary combustion chamber to a  minimum temperature
of 2,100ฐF. Future modifications may include conversion to fuel
oil. A photograph of the secondary combustion chamber  taken
during fabrication is shown in Figure 3.
                                                            Figure I
                                                       Process Flow Diagram
454    INCINERATION

-------
                           Figure 2
          Photograph of the Rotary Kiln During Fabrication
                          Figure 3
Photograph of the Secondary Combustion Chamber During Fabrication
  Overhead gases from the secondary combustion chamber are
cooled to l.OOOฐF by a water spray in the spray tower. Panicu-
late fallout is directed by a screw conveyor to the ash collection
system.
  Discharge gases from the spray tower are directed through two
heat exchangers that operate in series to reduce the temperature
of the flue gas to approximately 500 ฐF. The first heat exchanger
preheats combustion air from ambient conditions to 700 ฐF and
cools the  flue gas  to  approximately  SOOT. The second heat
exchanger (waste heat exchanger) further cools the flue gas down
to SOOT.  Paniculate fallout is conveyed to the ash collection
system.
  Exhaust gases enter the baghouse for paniculate removal. The
baghouse is designed to handle 22,000 actual  ftVmin of flue gas.
The maximum emission level is 0.08 gains/dsft3 (corrected to 12%
carbon dioxide). The baghouse is of the jet pulse design, i.e., high
pressure air periodically pulses to remove particulate that has
accumulated on the bags. Particulate is conveyed out of the bag-
house hopper and directed to the ash collection system. Gases dis-
charging from the  baghouse are directed to the scrubber for
HCl/acid gas conversion.
  The flue gas is scrubbed with a caustic solution to neutralize the
acidic gas component (HC1). The scrubber liquor (i.e., water and
salts generated upon combination of acid and caustic) is collected
for  on-site treatment and reuse. Scrubber blowdown water is
                                                                   filtered and used as system makeup water for ash cooling and dust
                                                                   control. Scrubber gases are directed to the stack for atmospheric
                                                                   discharge.
                                                                     Figures 4 and 5 are photographs of the installed TIS at the Lauder
                                                                   Salvage Yard site.
                                                                                             Figure 4
                                                                          Installed TIS at the Lauder Salvage Yard Site
                                              -View 1
                           Figure 5
        Installed TIS at the Lauder Salvage Yard Site
-View 2
MONITORING PROCEDURES
  A continuous emissions monitoring (CEM) system will extract
a sample of flue gas and analyze it for the following parameters:

• Carbon monoxide
• Carbon dioxide
• Oxygen
• Total hydrocarbons

SCHEDULE
  The initial schedule for activities at the Lauder Salvage Yard
is shown in Table 1. From contract award to commencement of
construction of the  TIS, the permitting activities constitute the
critical path. These activities are scheduled to be  complete in
6 months.
  This initial installation required 18 weeks for construction and
shakedown operations of the TIS and support equipment. The plan
is to study the erection, gain experience and make modifications
to the design and procedures to allow for completion of these
                                                                                                     INCINERATION     455

-------
activities in 7 weeks on future projects.
  The demonstration Test Burn (Trial Burn) is scheduled to be
executed in 1 week. This process demonstration will be conducted
while the TIS processes background soil spiked with PCB Aroclor
1260 to a concentration of 10,000 ppm (1% by weight).
  The duration of remedial operations will be dependent on the
amount of contaminated soil to be treated. It is expected that this
work will be completed in 12 weeks.

COMPLETION
  Site closure operations will begin at the conclusion  of all site
activities related to the TIS remediation program. Initially, site
closure may require the removal of the top 2 in. of crushed stone
which constitutes the surface layer of the hot zone site access road-
way. This material, which may be slightly contaminated,  will be
stockpiled in the soil staging area and gradually fed into the TIS
to achieve decontamination. A one-time pass-through of this
material should be sufficient to reduce the minimal contamination
that may be present. The treated material will be analyzed to ensure
that there is less than 2 ppm per resolvable chromatographic peak
in the PCB retention window. Upon receipt of the analytical results,
the crushed stone will be redistributed over the surface of the access
roadway.
  Upon completion of the crushed stone decontamination, the
remainder of site closure activities will begin. These activities will
include the removal, treatment and off-site disposal of the con-
crete decontamination pad and the HOPE liner in the soil staging
and processing areas. The concrete pad and liner will  be decon-
taminated by rinsing with high pressure water and, as required,
by detergent, surfactants and/or solvent rinse. The decontamina-
tion aqueous waste will be treated on-site (i.e., filtered and adsorbed
on granulated activated carbon). The spent carbon will  be treated
in the TIS. The decontaminated concrete and HOPE  cover will
be wipe tested to  ensure that the PCB concentration is less than
10 ug/100 cm2.
  When thermal treatment of contaminated waste is completed,
background soil will be processed in the system to purge the com-
ponents (especially the waste feed systems) of contaminated soil.
In addition, the internal and external portions of all process equip-
ment  (with  the exception of the  rotary  kiln  and secondary
combustion chamber internals) will be decontaminated using a high-
                            Table I
     Initial Schedule of Activities at the Lauder Salvage Yard Site
Activity
Contract Award
TSCA Pirnlttlnq
ICPA Permitting
TIS Fabrication
Sit* Preparation
Hobllliation
Refractory Cur*
Shekadow Operation*
Trial Burn Teatian;
Teit Report Approval
D*ปobl Illation
10/06/61
10/66
10/60
10/66
01/67
05/67
05/67
06/67
07/67
09/14/67
0ซ/21/67
0 1 /04 /H
04/01/66
04/21/66
• /A
04/67
04/67
04/67
04/67
05/67
07/67
06/67
0ซ/67
0k
] mo
75 A*Y1
11 dayi
21 dari
pressure water sprayer and detergent solution. Decontamination
waters will be treated on-site with granulated activated carbon.
Spent carbon will be incinerated in the TIS.  To ensure that the
equipment is properly decontaminated,  wipe samples will be
collected.
  After the conclusion of contaminated waste thermal treatment
at the Beardstown site, the waste feed system will be shut down
and the rotary kiln  and secondary combustion chamber will be
operated at their maximum design temperatures for a minimum
of 12 hr. The high temperatures in these units will be sufficient
to decontaminate the  unit internals.
  Remaining unburned waste residues generated from decontami-
nation or  cleanup activities will be sent off-site  for treatment.
  All components of the TIS, weigh station, fuel storage tank and
support facilities will be dismantled and removed from the site.
  Closure activities will include the final grading of the earthen
berms that surround each staging area to create a flat, gently sloping
surface. At the conclusion of closure operations, the site, with the
exception of the access roads which will remain, and several small
trees and  shrubs, which were removed,  wUl be restored to its
original state.
456     INCINERATION

-------
                    Automated Analysis  of  Adsorbent Traps  Used
                   In  the Volatile Organic Sampling  Train  (VOST)
                                                     Paul E.  Kester
                                                    Alan D.  Zaffiro
                                                   Tekmar Company
                                                    Cincinnati, Ohio
 ABSTRACT
   The analysis of emissions from hazardous waste incinerators is
 an important part of the regulatory permitting process. Destruc-
 tion and removal efficiencies (ORE) of volatile organic compounds
 must be achieved at levels of 99.99% or greater for a facility to
 obtain a permit. The Volatile Organic Sampling Train (VOST)
 protocol provides for sample collection and analysis of stack gases
 using solid sorbent traps. A ten position automatic sampler for
 analysis of VOST trap pairs has been developed and is described.
   VOST traps were spiked with a variety of compounds typically
 determined in incinerator stack gases. Data to be presented include
 recoveries of the volatile organic compounds from the traps and
 reproducibility across all 10  positions of the automatic sampler.

 INTRODUCTION
   With increased focus on alternatives to land disposal, hazardous
 waste disposal programs in  the United States are placing more
 emphasis on incineration. Environmentally, there is great concern
 that with incineration the waste problem becomes transformed from
 one of water and soil contamination to one of air contamination.
 As a result, incinerators which burn hazardous waste must be per-
 mitted prior to routine operation.
   Determining the concentration of volatile organic compounds
 in the stack is an important part of the permitting process. Incinera-
 tor efficiency must be  such that the  destruction and  removal
 efficiency (DRE)  of  principal  organic hazardous constituents
 POHCS are 99.99% or higher. Test burns usually are conducted
 to verify suitable operating conditions for a waste and determine
 the DRE.
   The Volatile Organic Sampling Train (VOST) was designed by
 the U.S. EPA to collect  stack gas samples on  solid sorbent traps
 for subsequent laboratory analysis. Because incinerator stack gas
 contains high humidity, VOST traps often contain large amounts
 of liquid  phase water in  addition to analytes. Water content of 3
 to 5 ml is not uncommon on a trap which contains 1.6 g Tenax
 TA. It is this water which presents  the greatest challenge to this
 analytical technique.
   VOST protocols are written as a part of S W-846, Test Methods
for Evaluating Solid Waste, Physical/Chemical Methods. Method
 0030 describes  the sampling train, while Method 5040 describes
 analytical procedures. In Method 5040, the two VOST traps utilized
 are analyzed as a pair. The primary trap contains Tenax and the
 secondary trap contains Tenax/Charcoal. An inert gas  flows
 through the traps while being heated to 190oC. Effluent from the
 trap travels through a heated line and enters the inlet side of a frit
 sparge tube of a purge  and trap concentrator which meets the
 specifications for Method 624, such as the Tekmar LSC-2, Model
 4000, or the new LSC 2000.
   The most common configuration to meet the requirements is the
 use of a clamshell heater with the concentrator. While effective,
 the method is very labor intensive. The analyst must be present
 at the beginning and end of each run to change samples. Sample
 throughput is also limited to the number of samples which can be
 run during working hours. Because all VOST traps must be blank-
 checked prior to being sent to a sampling site, additional analy-
 tical time is required.

DESCRIPTION
  The Tekmar Model 4210 Multiple Station Desorber was designed
to automate the analysis of VOST traps of the I/I design without
sacrificing the quality of the data. The Model 4210 has 10 indepen-
dent sample positions. Each sample position is equipped with its
own frit sparge tube for blank water, thus preventing carryover
from one sample to the next.  Furnaces are conveniently located
on the front panel of the instrument, and VOST trap pairs are
inserted from top to bottom and secured with 1/4-in. nuts and glass-
filled Teflon ferrules.
  Installation onto the concentrator,  which takes  only  about
15 min, is accomplished by making simple gas flow and electronic
connections. Connection of two  1/16-in. gas lines for desorption
and sample gases completes the  flow system.
  Electronic interfacing to the concentrator requires two cables.
The Prepurge cable carries the signal  to the concentrator to open
the purge valve for the Prepurge mode. Communication of initia-
tion of the purge step from the Model  4210 to the concentrator
and return of a purge complete signal to the Model 4210 is accom-
plished by the Concentrator cable. All electronic interfacing to the
GC or GC/MS for GC start and GC ready is handled through the
concentrator. GC run time generally determines the total analysis
time of the sample. Desorption of a trap pair occurs while a GC
run is in progress, with analytes stored on the concentrator's solid
sorbent trap awaiting analysis.  This overlap of functions increases
the efficiency of the GC by ensuring constant operation.
  The Model 4210 can be used for as few as one and as many as
10 samples. A First Cycle switch allows any position to become
the initial position. Subsequent positions are desorbed in order until
the position indicated by the Last Sample Switch has been analyzed.
Sample position is clearly indicated by a digital display.
  Full coverage mantle style heaters provide efficient desorption
of traps. Temperature range for the heaters is ambient to 250 ฐC.
To prevent condensation or adsorption of analytes once they have
been desorbed from a trap, all lines and valve are heated and are
                                                                                                INCINERATION    457

-------
continuously variable from 20 ฐC above ambient to 150ฐC. Tem-
peratures are set by adjusting screw pots on the front panel and
are read from the digital  temperature  readout display. A six-
position switch interconnected to the display determines whether
a setpoint or actual temperature of the sample, line or valve is
viewed.
  Samples are sealed in a  leakfree system once installed on the
instrument. One trap pair is placed in line with the concentrator
at a time through use of a 22-port air actuated valve. A supply
of 20-70 lb/in.2 air is required for actuation.
  Two  timed functions, set with thumbwheel switches, allow
prepurging and  preheating  the sample. Prepurge places a flow of
the inert desorption gas through the VOST traps prior to heating,
thus reducing the degradation of Tenax.  Artifact formation is
minimized and  trap lifetime  increased.
  Preheat allows the sample to heat prior to the purge mode,
ensuring proper sample temperature for the specified time period
according to protocol.  Reproducibility of all flow volumes and
sample temperatures are critical to obtaining reliable data.
  An option which simplifies the addition of blank water into the
sparge tubes is available. Two  port valves and sample needles added
to the top of the sparge tubes allows blank water to be added to
or withdrawn from the sparge tubes without disconnecting the
glassware.

OPERATION
  The variables (Prepurge  and Preheat times and  Sample, Line,
and Valve temperatures and  first and last sample) are set.
  VOST traps are connected with 1/4-in.  stainless steel ferrules
and glass-filled Teflon ferrules. Traps are joined in  such a manner
that the flow of desorption  gas back flushes the traps  ensuring
complete desorption. Trap pairs are inserted into the desorption
furnaces with the sampling inlet at  the top.
  Setting the AUTO/HOLD/STEP switch to the AUTO position
starts  the analytical process. Desorption gas flows from the bottom
to the top of the trap pair, sweeping the analytes to the sparge tube
containing organic-free water. Non-polar compounds are not
retained by the  packing. Water and polar  compounds remain in
the sparge tube. The humidity level of the  desorption gas stream
is decreased to the level common to purge and trap analysis of water
samples. When  loading  samples, the first  sample  can be started
before the others are loaded, reducing idle time of the GC.
  Analytes are retrapped on a  solid  sorbent trap  in the concen-
trator and remain there until the concentrator receives the GC ready
signal. Desorption of this trap to the GC with simultaneous GC
and data system start completes the sample introduction. Baking
the concentrator's trap after  desorption  minimizes the chance of
carryover from  a high concentration sample. When the trap has
cooled enough to accept the  next sample,  the Model 4210 cycles
to the next sample position and restarts the sequence. The entire
process repeats until the trap pair in the desorption furnace desig-
nated as the last sample has  been analyzed.

EXPERIMENTAL
  VOST traps were packed with 1.6 g Tenax TA, 35/60 mesh and
conditioned at 225 ฐC.
  Conditions on the Model 4210 were:  Desorption gas Helium,
Prepurge 0.5  min, Preheat 1.0 min, Desorb gas  flow rate 40.5
ml/min and Desorb temperature 180ฐC. LSC-2 conditions were:
Purge time 10.0 min, Dry Purge time 2.0 min, Desorb Preheat
175ฐC,  Desorb  180ฐC, Bake 225ฐC.
  A Hewlett-Packard 5890 GC/F1D was used. Chromatography
was performed on a 60-m x 0.75-mm VOCOL column (Supelco)
with a 1.5 um film thickness. The carrier gas was hydrogen at 10.5
ml/min.
  A gas phase standard was prepared in a static  dilution bottle
and maintained at 65 ฐC at all times. VOST trap pairs were spiked
by  injecting the gas standard into a  flowing stream of Nitrogen
connected to the trap inlet. To simulate high water content traps,
3 to 5 ml of organic free water were injected into the trap pair.
The amount of water added to the traps was random to further
simulate the variability of real samples.
  An initial study of on column reproducibility showed a "foRSD
of less than 4.2^o for six injections over a 3-day period.
  Two VOST trap desorption studies were performed on the Model
4210, one with dry spiked traps and another with high water con-
tent spiked traps. In both cases, the traps were spiked and loaded
across all 10  positions of the  instrument. The Model 4210 was
allowed to operate  unattended overnight, with data evaluation
performed  the next morning. Results appear in Table  1.
                           Table  1
                   Vosl Trap Reprodoclblllty
.COMPOUND | |)9 1
CMjCl, 1391
CHCl,
1,1.1-TCA-
Bซnlซnซ
PCZ*
o-Xylซnซ
2371
1411
1741
260 |
1761
On-Coluซn
Injection Dry VOST
1 USD 1 Bซcovซrv I
.49




.It
.90
.JO
.14
.29
10]
10*
9*.t
91 . 1
91.5
16.]
USD i
2.53 1
).J7 |
1.71 |
3.94 1
i.U 1
ปปt VOST
1 Recovซrv 1 BSD ,
104 .ซ
113
99.1
99.3
93. J
94.9
.20
.90
.33
.98
.09
•1,1,l-Trlchloroซthซnซ
  An additional study was performed to determine the applica-
bility of this instrument as an automatic sampler for water samples.
Frit sparge tubes were equipped  with  valves for easy sample
introduction.  Organic-free water was  spiked according to the
protocols for Method 624. Results in Table 2 show reproducibility
for five positions.
                           Table 2
                       Water Standards
COMPOUND 1
Chloroform |
Bซn(ซnซ |
Tซtrachloro*thซnป |
o-Xylซnซ |
ua/litar
50
50
50
50
1 I USD
1 4.1
1 2.9
1 S.ป
1 5.1
CONCLUSIONS
  The Tekmar Model 4210, when connected to the LSC-2 purge
and trap concentrator, provides reproducible analysis of volatile
organic compounds from VOST traps of the I/I design. Up to
10 samples can be run automatically with only a few minutes of
the analyst's time required for setup. Blank checking traps prior
to sampling can be performed with the same efficiency. Because
it runs unattended, the analyst  can spend time on other projects
during  the day  and continue with analyses overnight, thus
improving laboratory productivity.
  Samples and blank water are loaded  quickly and easily. All
parameters are set in advance,  ensuring reproducible conditions
from  sample  to  sample.  Discrete  glassware for each trap pair
reduces the possibility of carryover from a high concentration
sample.
  Water samples  can be analyzed according to U.S. EPA protocols
for water and wastewater. Soil samples, when analyzed by the
Medium Level  method  under the RCRA protocols,  can be
analyzed.
 458     INCINERATION

-------
                                 Use of Mobile Incineration  To
                                    Remediate the  Lenz Oil  Site

                                                   James F. Frank
                                                   Mary  E. Dinkel
                                        Division of Land Pollution Control
                                    Illinois Environmental Protection Agency
                                                 Springfield, Illinois
                                                    Desi M.  Chari
                                                Harza Environmental
                                                   Chicago, Illinois
INTRODUCTION
  This paper discusses the use of a modular mobile incineration
system to decontaminate through incineration 7,000 tons of soil,
the contents of 200 drums and several thousand gallons of organic
liquid and sludges. The principal contaminants are volatile and
semivolatile organics and chlorinated compounds. The cleanup pro-
gram is being carried out by  the Illinois Environmental Protec-
tion Agency (IEPA) as an immediate removal action. A description
of the history of the site, cleanup objective, incineration system,
mobilization,  design parameters, tanks removal and decontami-
nation, ambient air monitoring, soil excavation, vapor control, site
safety, community relations, trial burn plan, operations and cost
analysis follows.
  Lenz Oil Services is located in a valley surrounded by an expen-
sive residential neighborhood to the north and east of the facility.
Therefore, there is significant community interest and involvement
in the cleanup operation as well as concern about the impact from
air pollution  emission from  the incinerator operations.  Noise
pollution is also of concern to the nearby residents. lEPA's
activities regarding air pollution monitoring and noise reduction
measures are also briefly explained in this  paper.

HISTORY OF THE SITE
  Lenz Oil Services, Inc., located 30 miles  south of Chicago,
operated  for  20  years  as  an oil and solvent  storage/transfer
operation. The company obtained the necessary operating permit
from IEPA in 1981. However, due to numerous violations of the
permit conditions and mismanagement of hazardous waste, IEPA
through DuPage County Circuit Court ordered Lenz Oil to initiate
immediate cleanup action and to file a cleanup plan. Lenz Oil failed
to carry out major portions of the court order and in April, 1986,
the company filed for  bankruptcy. The company left behind 38
tanks and 200 drums containing 250,000 gal of liquid waste in its
2-acre site. Soil testing by IEPA showed heavy contamination with
various volatile and semi-volatile organic chemicals down to the
bedrock level.

THE REMEDIAL INVESTIGATION
  The IEPA tasked on of their multisite contractors to perform
an environmental assessment of the tanks and drum contents at
the Lenz site. The assessment entailed the sampling and volumetric
determination of liquids, sludges and solids found in the tanks and
drums.
  The analysis revealed high concentrations  of oils and petroleum
products  and chemical  constituents. Chlorinated  compounds
usually found in degreasing and paint solvents were detected as well.
  IEPA sampled the soil to determine the extent of contamina-
tion and the impact on the groundwater. Soil borings and subse-
quent monitor well installation revealed both soil contamination
to bedrock and contaminated groundwater. The depth to bedrock
varied from 6 to 14 ft throughout the affected portion of the Site.
  The contaminated groundwater flowing east off the site was
found to impact the residents well across the road from the site.
Some of the contaminants found on the Lenz site were detected
in the well water  in varying concentrations. At  least  one con-
taminant, benzene, was found to be in excess of the permissible
drinking water standard  during one  sampling episode. Conse-
quently, potable water will be provided to replace the well water
for five adjacent homes.
  Table 1 lists volatile and semi-volatile compounds detected in
soil samples collected at various depths throughout the site. Listed
along with these compounds are the highest detection  for each.
Concentrations are reported in mg/kg. Table 2 lists volatile organics
detected in perched groundwater collected on-site during various
sampling events. Concentrations are an average and are reported
in /tg/1.

CLEANUP OBJECTIVE
  Drums and tanks on-site were inventoried and sampled. Leaking
drums were overpacked  and labeled.  Analysis showed that  the
drums contained predominantly oils, solvents and tar waste. Tank
contents were contaminated with oils  and solvents.
  Several groundwater monitoring wells were installed to test
groundwater up to 38 ft. Monitoring  wells  were sampled in  the
summer of 1986, showed the presence of organic contamination
in all  monitoring wells.
  The IEPA, under its immediate removal program, contracted
with Environmental Systems Company, Little Rock, Arkansas, for
the Lenz Oil cleanup project. The initial contract involves destruc-
tion of hazardous waste stored on-site and decontamination of
7,000 tons of soil (about 60% of the soil on-site). The remaining
40% of the soil is not as heavily contaminated and will be  addressed
during the next phase of the cleanup operation along with ground-
water cleanup.
  The cleanup objective for the Lenz Oil site is to incinerate all
liquid wastes to greater than 99.99% DRE and  decontaminate
sludges and 7,000 tons of soil so that the residues from the inciner-
ation will be less than 5 mg/kg total hydrocarbons, provided that
the kiln waste feed contains less than  50,000 mg/kg total hydro-
carbons. If the excavated soil  exceeds 50,000 mg/kg,  it will be
blended in the feed area with other less contaminated soil in order
to reduce the feed soil to less than 50,000 /ig/kg total hydrocarbons.
                                                                                                 INCINERATION    459

-------
                                 Table 1
                   Lenz Oil Service Soil Sample Results
                    Reported In parts per billion 
337 organic compounds were detected  Including 4 carcinogens.   (average
               detected co"
carcinogenic compound Is 1610 ppb)
 concentration o
                   ected compound Is 30.362 ppb-ayerage concentration of
 NOTE:  Summary conttlni only Identified compounds detected  ปt concentritlons
 100 ppb.
 COMPOUND
 Acetone                         41.000                     4,937
 Benzene                          9.200                     1.575
 1.2 Olchlorobeniene              34,000                     	
 Ethyl Benzene                   520.000                    97.033
 Ethyl Mentyl Benzene             220.000                   125.500
 Dimethyl Benzene                680.000                   415,000
 Trlncthyl  Benzene               ISO.000                    83,500
 Chloroethane                       210                    	
 Dtchloroethane                   68.000                    IS.OB4
 1.1.1-Trlchloroethane             22.100                     3.413
 TrlcMortrtfluorethine             1,540                       268
 Tetrachlorethane                  1,300                    	
 TrIChloroethane                      60                        30
 Tetrachloroethene                 1.530                     3.442
 Trans 1,2-dlChloroenthene         80.000                    40.006
 Methyl Butane                      280                    	
 Heiane                             100                        25
 Pentane                           285                    	
 3-Methylptntane                    210                       100
 2,3.7-TMซethy1octane                170                    	
 3.7-Dtซethylnonane                 1200
 3-Methyloctane                     5270
 Decant                            820                    	
 2-Methyldecant                     180                    	
 Undecane                          1420                    	
 3,6-OlMthylundecane                2770                    	
 Mtthylene  Chloride               20,000                     1,422
 Toluene                        890,000                   142.413
 Xylenes                      2.000.000                   295.794
 Naphthalene                      30.000                    28.546
 2-Methylnaphalene                65.000                    21.060
 1-Methylnaphthalene               5.120                      2374
 Ethylnaphthalcnc                 15.000                    10.100
 DlKthylnaphthalene              47.000                    28,879
 Trlaethylnaphthalene              21.000                    11,941
 Decahydronaphthalene                610                    	
 Fluorene                         37,000                    12.555
 Bซnzo (a)  Fluorene                17.300
 Benzo (b>  Fluorene                4.570                    	
 Benzo Fluorene                    11.082                    	
 Pyrene                          83.000                    33,992
 Benzo (a)  Pyrene                 40.605                    22,969
 Benzo Pyrene                     20.122                    	
 Ctirysene                         46.201                    33.052
 Phenanthrene                    170.000                    64.697
 4-Methylphenanthrene              7,840                    	
 Benzo  Phenanthrene             9,810                    	
 Anthracene                       54.810                    42.270
 Benio (a) Anthracene              46.145                    25.960
 Dlbenzo (a.h) Anthracene           7.542                    	
 Fluoranthene                    200.000                    76.980
 Benzo (b) Fluoranthene             38,000                    22,411
 Benzo (k) Fluoranthene             27.000                    18.740
 Benzo  Fluoranthene             8,300                    	
 Benzo (g.h.U Fluoranthen*         8.300                    	
 Benzo (g.h.l) Perylene            20.000                    18.040
 Benzo Fluoranthene               25.893                    	
 Dtbenzothlspene                   7,400                    	
Carbazole                        10.700                    	
 2-Methlpropanol                   1.500                      610
 2-Propanone                        680                      438
 Butanolc add, methylester           190                    	
Ot-N-Butylphthalate               19,000                    5.698
Arochlor 1248                     9.000                    	
Arochlor 1260                     6,000                    	
 2-Heปanone                         260                      157
 2-8utanone                         860                      490
 4-Mcthyl-2-pentanone                230                      210
 Dlhydrofuranone                  25,953                    6.995
 5-Hethyl. 3-heซanone                692                      639
Dlbenzofuran                    27.538                    	
Adeno-1,2,3. Cd  Pyrene             4,166                    	
 2-Isopropyl-l,3-dlmethyl
 Cyclopentane                       160                    	
 l-Ethyl-2.2.6-trlmethyl
 Cyclohetane                       180                    	.
 Decahydro-2-methylnaphthalene        700                    	
Hexanedlolc acid, dloctylester      35,452                    5,849
Propanec add, ethenylester          100                       56
 Bis (2-ethylhexyl) phthalate      48.000                    17 355
 250 UNKNOHN OR UNIDENTIFIED HYDROCARBONS HERE DETECTED
 (average concentration of unknowns/unidentified   22,840)
                            Table 2
   Lenz Oil Service Summary Sheet Showing VOC Contamination
        Volatile Organici Detected in Perched Oroundwater

volatile organlcs detected In perched C.M.  Concentrations are averages.

                                                ua/l (PPb>

                                                    35

                                                    33

                                                    40

                                                    10

                                                    20

                                                   393

                                                   164

                                                    55

                                                   296

                                                   538

                                                  1200

                                                  6100

                                                    60
                                                                            Methylene Chloride

                                                                            I.l-Dlchloroethane

                                                                            Trans-l,2-Olchloroethylene

                                                                            I.l.l-Trlchloroethane

                                                                            Tricoloroethylent

                                                                            Benzene

                                                                            Toluene

                                                                            Ethylbenzene

                                                                            Xylene

                                                                            Acetone

                                                                            Aliphatic Ketones

                                                                            Aliphatic Alcohols

                                                                            Aliphatic Acid Esters

                                                                            Tetrahydrofuran

                                                                            Propylfuran

                                                                            Propyl Ether
                                                   322

                                                    30

                                                    65
                                                                          In the event the soil and sludges prior to incineration contain a
                                                                          level of contamination in  excess of 50,000 fig/kg, the soil and
                                                                          sludges after incineration are to contain the lesser of (1) 2 fig/kg
                                                                          or less of each organic priority pollutant and PCB or (2) a 10,000
                                                                          fold reduction for each organic priority pollutant and PCB. Any
                                                                          soil  and sludges not meeting the cleanup objective stated above
                                                                          will  be processed through the incineration system again until they
                                                                          meet the cleanup objective.

                                                                          SITE CLEANUP
                                                                          Tank Removal and Decontamination
                                                                            As part of ongoing operations at the  Lenz site, underground
                                                                          storage tanks and  upright storage tanks Were used to hold the
                                                                          liquids handled by Lenz Oil. The lack of structural integrity to the
                                                                          tanks and overall mishandling of liquids during the operations con-
                                                                          tributed in large pan to the contamination of soils.
                                                                            A  total of 16 tanks are to  be addressed during the mobile
                                                                          incineration remedial action. Of these 16 tanks, three underground
                                                                          tanks are  constructed of concrete  block  and   13  tanks  are
                                                                          constructed of steel plate and were semi-buried or sat totally above
                                                                          ground.
                                                                            All steel tanks were pumped out, decontaminated  using high
                                                                          pressure steam and cut up for scrap  resale. The three concrete
                                                                          storage tanks were used to store the liquids and sludges pumped
                                                                          out of the steel tanks. When all liquids  and sludges were removed,
                                                                          the concrete tanks were disassembled  and  the blocks crushed in
                                                                          place or placed in the shredder along with soils. If the blocks could
                                                                          not be passed through the shredder, they were staged for conven-
                                                                          tional disposal.
                                                                            The work in all of the tanks has been completed.  Five steel tanks
                                                                          required additional handling due to the construction of the tanks
                                                                          and the highly viscous residual material. The five steel tanks were
                                                                          cut lengthwise to enable a backhoe to access the residual material.
                                                                          These tanks were then decontaminated using high pressure steam
                                                                          and cut up for scrap resale.

                                                                          Soil Excavation
                                                                            The  extensive amount of soil  contamination  represents the
                                                                          principal reason for  implementing the mobile  incinerator  tech-
                                                                          nology.  Approximately 7000  tons of contaminated soil are to be
 460     INCINERATION

-------
excavated and incinerated. The excavation is performed by a track
mounted hoe which transfers the soil to a dump truck. The truck
transports the soil to the opening of a storage shelter where a front
end loader then transfers the soil to the opposite end of the storage
shelter. The shelter is used to prevent the soil from being exposed
to moisture and to contain any volatile emissions. A separate front
end loader transfers the soil to the shredder/feed hopper after being
weighed.
  Because of the possibility of high concentrations  of volatile
organic escaping to the atmosphere, a vapor suppressing foam was
applied during the initial excavation. The foam is used as a short-
term application during excavation and long-term application for
vapor suppression while being stored prior to incineration. During
the actual excavation, a short-term foam is applied at the mouth
of the excavation,  on the bucket of the end loader and on the dump
truck as it is  filled.  While the contaminated soil is stored in the
enclosed shelter, a  long-term foam is applied as needed  which
suppresses vapors up to 7 days. During normal operations, soil
excavation will be scheduled to occur once a week. This schedule
will provide adequate operating volume for one (1) week and help
minimize ambient emissions.
  Also during the  excavation, an HNu direct reading photoionizing
instrument is used to determine "fence line" concentrations above
background as a residential health and safety precaution. Sustained
excursions above  the predetermined target level of 5 units above
background as indicated by the monitor will be cause for modifi-
cation in excavation activities ranging from increased foam appli-
cation to ceasing operations until background levels are stabilized.

SITE SAFETY
  Safety and health at the Lenz Oil site have  been comprised of
two separate  plans: an on-site plan for the on-site Agency and
ENSCO personnel and an off-site plan for the residents within a
0.5-m radius  of the site and beyond.

On-Site Safety and  Health
  As a requirement  for every hazardous waste cleanup in the State
of Illinois, a Site Safety Plan must be developed and approved by
the IEPA before  on-site work can begin. The safety plan must
address level of protection relative to work performed; contain a
description of the on-site cleanup activities; show a site map to
clearly define  the hot zone, transition zone and clean zone; develop
decontamination  procedures for personnel and equipment; and
have a contingency plan for emergency incidents.
  The Lenz Oil site  is unique in that the cleanup and disposal will
be performed on site rather  than be removed for off site disposal.
As a result, the Site  Safety Plan is more comprehensive to address
both removal  and disposal than if the contaminated soil and liquids
were transported  off site.

Off-Site Safety and Health
  As part of the overall Site Safety Plan, the IEPA has  included
a comprehensive safety and health plan for all persons  living or
working within a 0.5-m radius of the site. Existing contingency
plans for incidents occurring at the  site have been supplemented
and broadened to encompass the nearby business and residential
community.
  The IEPA  has implemented preventive measures for residents
immediately adjacent to the site by relocating five households
during the initial excavation of soils, which has been determined
to be the time of highest exposure. During the first excavation
episode, monitoring  with direct reading instrumentation determined
that the need for continued relocation was not needed. Since the
need  for  future  relocation (during excavation  only)  was not
warranted, residential relocation would occur only in an imminently
dangerous situation. In the event of an incident requiring evacua-
tion, the IEPA has  arranged for a rapid alter notification system
which has the capability of notifying up to 400 households by phone
within 20 min. A total of 60 households or businesses are affected
in the 0.5-m radius. The system will be housed and operated by
the County Emergency Services and Disaster Agency (ESDA), the
principal contact and  coordinator of fire, police, American Red
Cross and other support agencies.
  The IEPA has additionally been participating in meetings with
all response agencies that would be involved in the event of an emer-
gency incident. As a result of the meetings, a finalized plan has
been generated and is in place for immediate implementation if
needed.

COMMUNITY RELATIONS
  Community relations involvement for the Lenz site has been very
high for many reasons. Chief among the reasons for this involve-
ment is the fact that  mobile incineration is a new approach to
remediating hazardous waste and the possibility of continuing oper-
ations beyond the predetermined volumes is of great concern to
the immediate and surrounding communities.
  Additionally, organized  community action groups which focus
on  environmental  concerns have rallied support  from the
surrounding communities to oppose mobile incineration. The com-
munities affected by the mobile incineration have a high aware-
ness  and interest  owing  to a  high  educational/professional
background. The high awareness of the mobile incinerator opera-
tion is additionally fostered by  health professionals and local
political  representatives.
  As a result of the high interest fueled by apprehension, the Lenz
Oil Review Committee was  formed. Political representatives, health
professionals and home and business owners meet monthly to
discuss their concerns about the operation  of a mobile incinerator
near their community.  The  concerns are responded to by the IEPA
toxicologists, air pollution permitting specialists,  noise pollution
specialists, an industrial hygiene specialist and the ENSCO mobile
incinerator experts.
  On-going site activities,  air emissions, noise from the ancillary
operations and contingency plans  in the event of a major system
upset make up the agenda for each  meeting. The meetings will con-
tinue as  long as there is community involvement and concern.

INCINERATOR MOBILIZATION
  The IEPA obtained exclusive possession of a 2-acre area adja-
cent to the Lenz Oil site for the installation  of the modular inciner-
ation system. After obtaining the necessary approval from IEPA
in June,  1987, ENSCO started to prepare the area for mobilizing
the units. Site preparation included installation of a sewer and water
line, grading, and grading with sand and gravel. ENSCO installed
a work trailer, decontamination  trailer,  including lockers and
showers  and three  office trailers.
  The mobile incineration  system addressed in this paper consists
of several pieces of equipment on separate trailers. This system
is trailer and skid mounted  and several units are pre-piped and pre-
wired to a certain extent. However, it needs assembly on-site before
operations commence and  require disassembly prior to transpor-
tation to another site.
  On June 9,  1987, ENSCO delivered to  the site 20 truck loads
equipment that comprised of the modular incineration system con-
sisting of rotary kiln, secondary combustion chamber, cyclone sepa-
rators,  boiler, air  pollution  control  scrubber  system, water
treatment units, steam ejector, demister, stack assembly  and the
master control trailer. Most  of the large equipment was built on
to the trailer bed and other assemblies were skid mounted and could
be moved to various locations on-site.
  ENSCO started  to  assemble all equipment on June 10 using
cranes. The construction was completed on July 20,1987. The kiln
was relined with refractory bricks and the system was ready for
shake down and startup. The system was initially started up by
burning natural gas. All equipment was tested and the refractory
in the kiln was cured over a period of 48 hr to reach a steady state
condition. ENSCO was ready for trial burns on July 25, 1987.
                                                                                                       INCINERATION    461

-------
                                                          Figure 1
                                            ENSCO Environmental Services Site Layout
AMBIENT AIR MONITORING
  Inasmuch as the Lenz Oil site is located near residential areas,
the public was very much concerned about the impact of air
pollution from the operation of the incineration system during the
cleanup. IEPA installed an air monitoring system around the site
to determine the impact of volatile organic compound emissions
from the cleanup operation. Initially, IEPA installed two ambient
volatile organic compound emission  monitoring  stations, one
upwind and another downwind of the facility to determine the
background level prior to the startup  of the operation.
  Six 24-hour samples were taken from each site prior to the com-
mencement of the operation. The sample canisters were sent to
Radian Corporation, in North Carolina for analysis. The samples
were analyzed for total non-methane hydrocarbons.  Should the
        ?**•ซซ ซซT ซOU
                           TCMURMW
                                 l/UOUDi HCMIUTO* THM.ni
results show ambient levels higher than 5 ppm, specific compounds
analysis will be performed on the duplicate sample.
  IEPA has located three additional ambient air monitoring
stations around the Lenz Oil site. The Agency has contracted with
private consulting company, to run  these five air monitoring
stations for the duration of the cleanup operation. A minimum
of two 24-hour samples per week will be collected and analyzed
for total non-methane hydrocarbons. Samples will be analyzed for
selected organic compounds in some situations.
                          Figure 2
                    Equipment arrangement
462    INCINERATION

-------
PROCESS DESCRIPTION
  The process  flow and equipment arrangement are shown in
Figures 1, 2 and 3. The major components of ENSCO's mobile
incineration system MWP 2000 are discussed below.

Waste Feed System
  The solid waste feed system consists of a hopper, shredder,
conveyor, feed hopper  and screw auger. At the  Lenz Oil site,
excavated soil is stored inside a covered tent area. Soil from the
tent area is loaded into the hopper after its weight is recorded. This
hopper is attached to a shredder where large materials are reduced
to 4 in or less. The shredded materials are then transported by an
enclosed belt conveyor to the 2-yd3 feed hopper. The feed hopper
is totally enclosed and has vibrating blades for easy feeding of wet-
sticky materials. This hopper feeds the soil onto the auger which
is a 12-in diameter screw. The auger is inclined into the kiln at 30
degrees from horizontal. The feed auger has a variable speed con-
trol which is used to determine the feed rate to the kiln.
  The contaminated liquid feed system consists of a flat straight
nozzle which introduces liquid as a horizontal spray into the kiln
so that it intersects the burner flame which is in the combustion
zone of the kiln. The waste feed system consists of a pair of pumps
and strainers. An automatic shut-off valve is attached to the auger
system for waste feed cutoff.
  The kiln also is equipped with a sludge injection nozzle which
could spray pumpable sludges into the kiln. The nozzle is supplied
with steam to atomize the sludge being sprayed into the kiln.

Rotary Kiln Incinerator
  The rotary kiln is a carbon steel cylinder mounted horizontally
on a custom trailer and is lined with 6-in of refractory brick. The
kiln is 30 ft long and has a 7.6 ft OD and a 5.5 ft ID with an effec-
tive volume of 697 ft3. The kiln is mounted so that it can be tilted
up to six degrees with variable speed of one half  to 4 rpm. The
burner and the solid liquid waste feed system are located  at the
higher end of the kiln while the gas outlet and ash drop are located
at the opposite lower end.
  The kiln burner is designed to provide a thermal loading up to
14 million Btu/hr and outlet gas temperatures of 1,400—1,800
degrees F. The kiln will be operated at stoichiometry of 1.1—1.5.
The solid residence time which will vary with the rate and type of
waste being fed and can be varied by changing rotation speed of
the kiln, is anticipated to be 30 to 60 min.
  Ash and decontaminated soil from the rotary kiln are discharged
into a field breaching at the lower end of the kiln. These materials
fall from the breaching into an ash receiving tank that is filled with
water above the discharge lip of the breaching to provide a water
seal.
  The hot gases from the kiln pass through a pair  of cyclones,
installed in parallel, before entering the secondary combustor. The
cyclones remove particulates entrained in the gas stream. They are
lined with 4-in. of castable refactory.
  Particulates removed from the gases by the cyclones flow by
gravity through tubes into the ash receiving tank. The water con-
tained within the ash removal unit quenches the hot particulate
solids and serves as a liquid seal between the kiln, cyclone duct
work and  transition  duct work to the secondary combustor.
  The ash collected is removed using a chain drag conveyor and
is discharged into a metal bin to be removed using  a bobcat. The
treated soil is then stored on the site until the laboratory results
show it is  clean. Then it is placed  back  in the excavated hole.

Secondary Combustor
  The  secondary combustion chamber  (SCC) is  designed to
incinerate  the volatile evaporated  in  the rotary kiln and oxidize
the compounds to HCL, H20, and CO2. The SCC has a 24 million
Btu/hr thermal loading range and will reach a maximum tempera-
ture of 2.400 degree F. Typical stoichiometry will be 1.2—1.5.
  The SCC is a carbon steel cylinder mounted horizontally on two
supports on a custom trailer. It is lined with 2.25 in of insulating
brick and 4.5 in of fire brick. The SCC is 40 ft long with an inter-
nal diameter of 6.5 ft. The effective volume of the SCC is 1377
ft.3 with an expected residence time of 2 sec.

Waste Heat Boiler and Auxiliary Components
  Hot gases from the SCC pass through the waste heat boiler to
recover  the heat. The boiler can use steam at 250 lb/in.2 which
is supplied to the ejector scrubber and deaerator. The boiler is rated
at 19.6 million Btu/hr. The boiler is designed to maintain a tem-
perature of 400 degree F to avoid corrosion, and it is also designed
to maintain high velocity through the tubes to avoid particulate
deposition. The gases entering the boiler around 1600-1800 degrees
F exit the boiler around 400-600 degrees F after heat recovery.
  Steam produced by the boiler is supplied from the steam drum
to its several uses through the system header. Water used in the
boiler is treated to avoid chemicals deposits  in the  boiler.

Air Pollution Control Units
  The air pollution control system consists of a quench elbow,
packed tower, ejector scrubber and packed effluent neutralization
unit. This system is designed to cool and remove acid and sub-
micron size particulates from the gases that exit the waste heat boiler
and neutralize the effluent generated in  this system. The quench
elbow contains several nozzles which spray recirculated water from
the effluent neutralization tank into the elbow to cool and partially
remove  acid from the gases that exit from the waste heat boiler.
  Gas temperatures are reduced from around 500 degrees  F to
approximately 165 degrees F. The quench elbow is made of Inconel
which has an initial inside diameter of 4.5 ft and a final diameter
of 2.5 ft at the inlet into the pack tower inlet duct.
  The neutralization tank collects the recirculated water sprayed
into the quench elbow  and the water drained from the packed
tower. Raw makeup water and caustic will be added to the system
to maintain a pH of approximately 5.

Packed  Tower
  The packed tower is designed to remove additional acid from
the gases that exit the quench systems.  The gases flow upward
through the tower and are scrubbed by a countercurrent flow of
water sprayed into the top of the tower. Scrubbing water is in-
troduced through individual spray nozzles located on the top of
the tower. Water from the tower flows by gravity into the effluent
neutralization tank.
  The packed tower which is 14 ft high and 6 ft in diameter and
is  fabricated of fiberglass reinforced  plastic. It is filled  with
approximately 6 ft of 2-in. diameter plastic saddle packing material.
A demister pad is installed above the packing.

Ejector  Scrubber, Demister and Stack
  The ejector scrubber  is designed to  remove submicron size
particulate and additional acid from the gases before they are dis-
charged through the demister and the stack. Gases from the packed
tower are drawn through the ejector mixing tube by the force of
steam delivered through the nozzle in the mixing tube. The high
turbulence created by the steam nozzle system causes agglomera-
tion of submicron particles and the absorption of acid in the water
vapor supplied  by the steam. This material is removed by the
removal of water vapor in the demister at the downstream end of
the scrubber.
  The ejector scrubber also serves as a prime mover for the entire
system.  The drawing of gases through  the ejector mixing tube
produces up to 25 in. of water column vacuum. This is sufficient
vacuum to draw gases through the entire system and eliminate the
need for an induced draft fan.

Stack
  The stack is fabricated of fiberglass reinforced plastic and is
35 ft from the trailer bed and approximately 40 ft from the ground
level. The stack is equipped with sampling ports and an access plat-
                                                                                                      INCINERATION    463

-------
form to these ports to facilitate sampling during trial burns. The
stack is equipped with the gas sampling system that collects, con-
ditions and delivers the continuous stack sampled stream to O2,
CO2,  CO2 and HC analyzers located in the control room.

Trial Burn
  Under the  RCRA  regulations, all incineration systems  must
demonstrate that they can achieve: (1) a destruction and removal
efficiency (DRE) greater than 99.99%.  (2) paniculate emissions
less than 0.08 grains/dft3 and (3), Hcl control efficiency greater
than 99%. Although ENSCO does not require a Part B permit from
IEPA. ENSCO still must meet all the RCRA requirements. Trial
burn  results  will  provide performance  data  and  operating
parameters for  the incineration system.
  Lenz Oil site is contaminated with various hazardous constituents
at different concentration levels. Therefore, ENSCO has proposed
to introduce known quantities of specific POHCS with the soil and
liquid wastes to demonstrate the performance of the incineration
system during the trial burns.
  The test burn program  consists of three tests, each with three
replicate performances. The three tests will be conducted at soil
feed rates of 8000, 10,000 and 12.000 Ib/hr. The liquid feed rate
will be 1000 Ib/hr for all tests.
  The trial burns will involve spiking the contaminated soil and
liquid with known quantities of POHCs in order to demonstrate
DREs. ENSCO selected carbon tetrachloride (CCL4) and Perch-
loroethylene (PCE) as the POHCs. CCL4 was selected as a POHC
due to its low heat of combustion (0.24 kcal/kg) and PCE was
selected because it has similar characteristics to some of the con-
stituents found in the soil samples at the site. Successful incinera-
tion of these two compounds with a DRE greater than 99.99"% will
demonstrate that the ENSCO mobile waste incineration system
(MWP2000) can  meet regulatory  performance  standards for
halogenated and non halogenated hydrocarbons.
  The excavated and stockpiled contaminated soil at the Lenz Oil
site will be spiked with a precise quantity of PCE so that a nominal
POHC concentration of 1 % by weight is achieved. This prepara-
tion of the feed will be done by dropping pre-weighed plastic bottles
filled with PCE into the feed anger at regular time intervals. To
maintain a PCE concentration of 1% of the soil feed, one bottle
containing 1.3 Ib of PCE will be added to the feed anger every
minute at a soil  feed rate of 4000 Ib/hr. Carbon tetrachloride will
be added to the waste liquid stream at a rate of 0.33 Ib/min, using
a posite displacement metering pump, to maintain the CCL4 con-
centration of 2% of the waste liquid  feed rate of 1.000 Ib/hr for
each test.
  ENSCO has contracted  Versar, Inc., Springfield, Virginia to
conduct the trial burns. Versar will be taking stack samples using
(1) volatile organic sampling train (VOST) for the hydrocarbon
analysis and (2) Method-5 sampling train for paniculate, HCL and
metals analyses. Stack gas samples will be through two 4-in. ports
on the stack which are located 90 degrees apart approximately
6 ft from the top of the stack. The VOST train consists of a Tenax
and a Tenax-charcoal cartridges  for collection of organic com-
pounds by adsorption. A total of 4 pairs of Tenax, Tenax-charcoal
samples will be collected during 80-min test run. and each pair will
be used for 20 min at a flow rate of 1  1/min through the VOST
system.
  The Method-5 sampling train will be used to collect samples from
six traverse points in each part for a period of 8 min at each traverse
point. The sample will be collected isokinetically for a total of 96
min/run.
  During the trial burns, Versar will collect the following samples
for the nine runs:

• Neat PCE and  CCL4
• Solid feedstock background  matrix
• Treated soil residues
• Stack gas (VOST and Method 5) Effluent neutralization tank
  background
• Effluent neutralization tank (after each run)
• Scrubber water sump
• Blanks

  All samples collected during the trial burns will be analyzed at
Versar's laboratory in Springfield, Virginia.

OPERATIONS
  After the completion of the trial burns, ENSCO will be allowed
to operate provided the  preliminary results from the trial burns
demonstrate (1) paniculate emissions of 0.08 grain/ft3 or less, (2)
HCL removal greater than 99% and (3) combustion efficiency ( =
100% x CO2/COj2 + CO) greater than 99%. The soil feed rate
during this interim operation will be limited to 4 tons/hr ENSCO
will be allowed to operate at the maximum rate of 6 tons/hr if
they meet the DRE of greater than 99.99%.
  Each batch of decontaminated soil be analyzed to verify whether
it meets the cleanup objective. After successful decontamination,
soil will be backfilled at the Lenz Oil site. The cleaned soil and
contaminated soil will be segregated using PVC liners to prevent
cross-contamination. Water accumulated in the excavated area due
to rain or groundwater recharge will be tested to determine the con-
tamination level. Based on the contamination level, the water will
be treated on-site or sent to the DuPage County water treatment
plant for treatment. Cost data for the operation are shown in Table
1.

           Cost Estimates of Treatment Prices
Initial
Activities

Treatment of
Contaminated
Soil and Sludges

Treatment of
Contaminated
Waste in Drums

Treatment of
Contaminated
Liquids in
Tanks

Treatment of
Contaminated
Soils and
Sludges
Quantity    Unit

             One


   7,000     Ton
  Unit
  Price     Extension

             $98,000


 $516.20   $3,614,030
     200   Drums    $400.00      $80,000
 150.000     Gals
            Tons
   $0.50     $75,000
Required
   Setup
$982,000
1,000
2,000
3,000
4,000
5,000
6,000
7,000
Tons
Tons
Tons
Tons
Tons
Tons
Tons
11.358.00
$867.00
$703.33
$621.50
$572.40
$539.67
$516.29
$1,358,000
$1.734,000
$2,109,990
$2,486,000
$2,862,000
$3,238,020
$3,614,030
CONCLUSION
  The IEPA has embarked upon an innovative technology to
replace the conventional relocation of hazardous wastes. Illinois
has identified over 300 uncontrolled hazardous waste sites, the
majority of which  are  contaminated with organic chemicals.
Thermal destruction offers a permanent resolution to the remedi-
ation of uncontrolled hazardous waste sites. The experience gleaned
from  the successful completion  of the Lenz Oil cleanup will be
instrumental in the remediation of these sites.
464    INCINERATION

-------
                      Development  of  a Gaussian  Puff Model  for
                            Over-Ocean  Incineration  Applications

                                                James G. Droppo, Jr.
                                                  Richard M. Ecker
                                     Battelle, Pacific Northwest Laboratories
                                                Richland, Washington
                                                    David Redford
                                      U.S. Environmental Protection Agency
                                                  Washington, D.C.
ABSTRACT
  Atmospheric emissions will occur from at-sea incineration of
hazardous wastes. A Gaussian puff model is described that can
predict airborne concentrations and  surface deposition rates
resulting from such over-water atmospheric emissions. MESOSEA
was developed from a land-based model for application to atmos-
pheric transport, diffusion and deposition processes over open
water surfaces such as the ocean. The model treats both precipi-
tation or nonprecipitation conditions over water surfaces. Written
in FORTRAN, the current version is implemented to operate on
an IBM-PC or compatible MS-DOS desktop computer.
  Model  routines were implemented for diffusion parameters,
deposition processes and wind shear (wind variation with height)
over water surfaces. Over-water dispersion rates are expressed as
a function of distance  traveled with coefficients that vary with
atmospheric stability classes. Stability classes are defined in terms
of wind speed and air-sea temperature differences. The dry depo-
sition processes and wind variation with height are implemented
in MESOSEA as a function of local surface roughness and ambient
atmospheric turbulence. The surface roughness is expressed as a
function of wind speed over the water surface.
  Outputs from test runs for a hypothetical burn operation of the
MESOSEA model are presented. This illustration is based on
meteorological observations taken at a proposed incineration site.
Plots of cumulative deposition on the ocean are shown to demon-
strate the utility of the model. The MESOSEA model can provide
computations of potential plume behavior for sites where sets of
real or forecast data are available. Although the model has some
limitations, it is appropriate for case studies of air concentrations
and deposition rates from at-sea incineration operation emissions.

INTRODUCTION
  The at-sea incineration of hazardous waste is receiving serious
consideration as a viable disposal option1. This process requires
evaluation of the proposed incineration sites for potential environ-
mental impacts. Modeling of material movement in the atmospheric
plumes from incinerator ships  must incorporate special features
of over-ocean processes. A model suitable for computations of un-
steady atmospheric plume characteristics and cumulative oceanic
deposition rates is described in this paper.
  The atmospheric emissions from at-sea incineration operations
are of concern both in  terms of atmospheric concentrations and
ocean deposition. The emissions will be affected by an atmospheric
marine environment; the proposed incineration sites are far enough
from shore that the releases should not be initially influenced by
shoreline  atmospheric dispersion processes.
  Potential emissions include trace quantities of a variety of com-
plex, organic compounds, trace metals and HC1. The actual in-
cineration emissions will depend on the nature and composition
of the feed material ship operating procedures and incinerator
design and operation. A variety of hazardous materials have been
proposed as feed material. Proposed designs have releases that
range from a very hot, vertical stack release to a relatively cool,
horizontal release.
  This paper describes the development of a Gaussian puff model
for atmospheric transport, dispersion and deposition computations
over water surfaces. The main advantage of puff models (i.e.,
MESOI2, MESOSEA and INPUFF3) over straightline Gaussian
models (i.e., PAL4 and ISC5) is their ability to account for spatial
and temporal changes in the wind fields. The model described in
this paper is a research model and not a regulatory model. An at-
sea screening model with less stringent data input requirements is
also under development as a possible regulatory model for at-sea
incineration operations6.

APPROACH
  Marine atmosphere processes will control movement and dilu-
tion rates of the airborne materials. Certain models developed for
overland surface application are inappropriate for at-sea applica-
tion. The land-water  differences in atmospheric processes are
mainly the result of changes in the surface energy budget and
surface roughness. The fixed, sometimes flexible, nature of the land
surface elements contrasts with the water surface, which can move
vertically and horizontally. The range in surface temperatures and
sharp surface thermal gradients tends to be considerably smaller
over water than over land surfaces. A water surface is relatively
smooth compared to most land surfaces, and the roughness of the
water surface increases with wind speed in a manner not observed
for land surfaces.
  These land-sea differences point out reasons why a land-based
model may not be applicable to at-sea computations over water
surfaces. First, dispersion coefficients are developed based on over-
land studies and are expressed in terms of overland parameters.
Second, computations within the code that depend on local surface
roughness and atmospheric turbulence may not satisfactorily
account for at-sea processes. More specifically, deposition rates,
plume rise and wind variation with height will be affected by the
smoother-moving surface.
  An at-sea Gaussian puff model, MESOSEA, was developed by
modifying an overland model, MESOI Version 2.02, for appli-
cation over open-water surfaces. MESOI Version 2.0 is a Gaussian
puff model designed to predict airborne and surface exposures

                                  INCINERATION     465

-------
resulting from releases from land-based  facilities.  MESOSEA is
a smaller, more efficient version of this code incorporating some
at-sea routines. Because MESOI Version 2.0 is a well-documented
code2, this paper will provide an overview of the model  and
documentation on areas where code modifications were made for
at-sea applications.

MODEL OVERVIEW
  MESOSEA and  MESOI Version  2.02  are third-generation
Lagrangian puff transport and diffusion models with direct lineage
to  a model  written  by Start and  Wendell7. These  current-
generation codes are interactive models developed to estimate air-
borne concentrations and surface deposition rates of contaminants
released to the atmosphere. Plume rise, atmospheric transport and
dispersion, wet and dry deposition and contaminant decay processes
are included.
  A Gaussian puff approach to modeling  atmospheric diffusion
assumes that material released to the atmosphere is distributed in
a Gaussian form  in  one or  more puffs. The plume from a
continuous release is  approximated by a series of puff releases.
The airborne concentrations and surface deposition rates are com-
puted by summing  the contributions of puffs traveling over the
location of interest. The concentration from a single puff without
the effects of surface and inversion layer  reflections and  plume
decay processes at a point in the atmosphere in  a Cartesian
coordinate system (x, y, z) is  given by:
  CHI (x,y,z)   = airborne concentration at point (x,y,z), g/m3

  Q            = mass of material contained in the puff, g

  Xo, y,>, and Zo = position of the center of the puff,  m

  ox, ay, and az = Gaussian diffusion coefficients, m

  C            = constant equal to (2ir)3/2

  At the time a puff is released, the center is assigned coordinates
(xo yo zo) where xo and yo are the horizontal coordinates of the
release point, and zo is the effective release height (actual release
height plus plume rise). The values of xo and yo change as a puff
is advected by the wind, but the value of zo remains the same.
  The dispersion coefficients are expressed as functions  of the
distance the puff has traveled and stability class. The distance the
puff travels is the sealer sum  of the path of the  puff.  The puff
speed is assumed to be the wind speed at the height of the puff
center.  Stability  classes  such as  those  originally  defined  by
Pasquill8  are  discrete subdivisions  of the range of possible
dispersion conditions. Definitions of stability class normally are
expressed in terms of indirect indicators of turbulence (i.e., solar
insolation, mean wind speed,  vertical temperature gradient); the
over-water dispersion approach used  in the MESOSEA model is
described below.
  Equation (1) provides an expression for the concentrations from
a puff assuming it is free  to spread to infinity in all directions.
In reality, the puff dispersion is limited vertically by the ground
surface and inversion layers in the atmosphere. MESOSEA retains
the multiple reflection factor formulation documented by Ramsdell
et al2.
  The puff concentrations are also corrected for contaminant losses
from radioactive or chemical decay, dry deposition and  wet depo-
sition. The material  loss is treated as a source depletion model;
depletion is assumed to occur uniformly throughout the puff. With
the exception of changes in the dry deposition computation noted
below, the removal and puff decay formulations are given in detail
by Ramsdell, et al.2.
  MESOSEA requires data files of hourly meteorological obser-
vations of surface wind speed and direction, upper-air wind speed
and direction, air temperature,  atmospheric stability category
(dispersion parameter) and precipitation parameters (duration and
intensity of precipitation). The user interactively defines various
run-time parameters (emission rate,  height, temperature, exit
velocity, grid size, contaminant deposition properties, etc.).

MESOSEA DEVELOPMENT
  Development of the MESOSEA code included changes in defi-
nition of input parameters and computation routines. The changes
made in the code are discussed below, along with the rationale for
the changes.

Atmospheric Stability
  MESOI Version 2.0 uses stability classes to characterize the pos-
sible range of dispersion rates for the atmospheric plume. Stability
classes are given letters A to G: A, B and C, are unstable; D is
neutral; and E, F and G  are stable conditions. Applying these
stability classes requires  the adoption of a revised method of
defining the stability classes for over-water conditions.
  MESOSEA defines the stability classes in terms of the wind speed
and air/sea temperature differences, using relationships defined
by Hasse  and Weber*.  The definition of stability category in
MESOSEA requires either quarter-hourly or hourly sea-surface
temperature, air temperature and wind speed data.

Wind Field Computation
  MESOI Version 2.0 accepts meteorological inputs at multiple
stations to account for spatial variations caused by topographical
effects. The number of stations was reduced to one in MESOSEA,
because topographical effect routines are inappropriate for at-sea
computations, and normally the most that can be expected to be
available is the data from a single meteorological observation site.
Hence, the wind field is computed in MESOSEA using winds from
a single observation point, allowing for temporal wind changes
while assuming spatial uniformity. In addition to being a reasonable
approach for over-water applications, the reduction in the number
of wind stations greatly increases the performance of the model
by greatly decreasing the memory requirements and the run-time
computations. This modification has helped allow a successful
implementation of MESOSEA on a personal computer.

Surface Roughness
  A routine to compute surface roughness as a function of wind
speed was incorporated to MESOSEA. This routine is based on
Charnock's relationship as described  by  Joffre10:

  Zo  = m u*2/g

where:

  Zo  = roughness length of surface, m

  m  = coefficient (  = 0.0144 recommended by Garratt<">]

  u*  = acceleration of gravity,  m/sec*


Wind Shear
  The vertical variation of winds is called the  wind shear. The
accuracy  of the  computation of the path of the puff directly
depends on how accurately the speed and direction of movement
of the puff are known. Quite different rates of wind shear are
observed in overland and over-water surfaces. Differences in wind
shear are  mainly the result of differences in surface roughness.
   MESOI Version 2.0 requires  input of wind  values from the
surface to the height of the top of the mixed layer in the atmos-
phere. The requirement for inputs of temporal definition of wind
466    INCINERATION

-------
profiles is difficult to meet even for land-based stations, let alone
the proposed remote incineration sites.
  MESOSEA gives the choice of using either an upper-level wind
speed or automatic internal computation of the upper-level winds.
The vertical  variation of wind  speed algorithm is  based  on
micrometeorological similarity relationships12'13. These routines,
which compute wind shear as a function of the roughness of the
sea surface and ambient atmospheric turbulence intensity, should
provide reasonable estimates of vertical wind speeds within  100
to 200 m of the  surface.  Sea surface roughness is computed as
described above. Atmospheric turbulence intensity is defined as
a preset value of Monin-Obukhov length for each stability class
as shown in Table 1. The friction velocity, a measure of atmospheric
turbulence intensity, is computed with an iterative routine based
on the wind profile similarity relationships using the known wind
speed, surface roughness length and Monin-Obukhov length. Once
a consistent roughness length and friction velocity are defined, the
similarity relationships are used to compute the wind variation with
height.
                          Table 1
     Summary of Approximate Central Monin-Obukhov Values
            For Each of the Pasquill Stability Classes
               (Derived from Hasse and Weber9)
        Pasquill
        Stability
          Classes

            A
            B
            C
            D
            E
            F
            G
Central  Inverse
 Monin-Obukhov
Length,  1/L,  1/m

       -0.60
       -0.28
       -0.03
        0.00
        0.12
        0.30
        0.50
Dry Deposition
  MESOI Version 2.0 uses a "deposition velocity," which is the
basis for computing the dry deposition flux. Because deposition
velocity is expected to vary greatly with both ambient meteoro-
logical conditions and properties of the material of interest routines
have been incorporated into MESOSEA that provide an estimate
of particulate or gaseous deposition velocities based on the ambient
sea-surface conditions (local roughness), stability and wind speed.
  For particles, the user needs only to supply a gravitational settling
velocity for the material of interest, and the model computes a total
deposition velocity using empirical deposition relationships14. For
gases, a "surface deposition velocity" must be approximated using
Henry's Law. The surface deposition velocity is equivalent to the
inverse of the surface deposition resistance. Lacking specific data
on  properties of a material,  a maximum deposition rate can be
obtained by assuming a zero surface resistance and by having only
the atmospheric turbulence over the ocean surface limit the rate
of dry deposition.
  The dry deposition routines  have been derived based on a
resistance approach to over-water deposition.  The relationships
used to compute the atmospheric resistance for dry deposition are
based on micrometeorological similarity equations12'13 assuming
an analogy with atmospheric heat  flux. Similar  approaches are
described in more detail in Van  Voris, et al.15 and in Droppo, et
al.ซ

APPLICATION
  To illustrate the capabilities of MESOSEA, test runs were made
for  several nonprecipitation conditions based on a unit emission
                                     rate from the incinerator ship. Values used for the incinerator stack
                                     characteristics (Table 2) were based on single stack emissions from
                                     the Vulcanus  II1. Meteorological data for these test runs were
                                     based on information in the ship's log from a previous baseline
                                     cruise at the North Atlantic site under late-fall conditions. The input
                                     data  for Nov. 12, 1985, as used in the test runs,  are given in
                                     Table 3. Neutral atmospheric stability and a 200-m atmospheric
                                     mixing  height were assumed for these test runs. The actual  sta-
                                     bility for the time period selected for the simulation was not known
                                     because the water and air temperature differences were not avail-
                                     able.  However, measurements made on other days during this cruise
                                     suggest that the stability was either C or D.
                                                               Table 2
                                           Incinerator Ship Characteristics Assumed for Test Cases
                                             Parameter
                                     Stack  exit  diameter
                                     Stack  exit  temperature
                                     Stack  exit  velocity
                                     Stack  exit  height
                                     Number of stacks
                                Value

                                   3.2
                               1400.
                                 15.
                                 12.
                                   1
Units

  m
  deg
  m/s
  m
                     Table 3
        Meteorological Data Used in Test Cases

                           Mixing   Wind     Wind
               Stability   Height, Speed,  Direction,
Date    Hour     Class        m      m/s     degree
                            200
                            200
                            200
                            200
                            200
                            200
                            200
                            200
                            200
                            200
                            200
                            200
                            200
                            200
                            200
                            200
                            200
                            200
                            200
                            200
                                       MESOSEA provides output options at 15-min intervals. Out-
                                    put options include listings/plots of undepleted air concentration
                                    (based on dispersion with no deposition), undepleted cumulative
                                    air exposure, present depleted air concentration (based on disper-
                                    sion with deposition), depleted cumulative air exposure, present
                                    ocean deposition and cumulative ocean deposition. In addition,
                                    MESOSEA  can provide similar outputs for a component being
                                    produced in  the plume by a chemical reaction or radioactive decay.
                                    Although each of these outputs may be of interest in certain appli-
                                    cations, the  cumulative deposition to the ocean was selected for
                                    use in the test  runs because of interest in incineration-at-sea
                                    applications.
                                       Cumulative ocean deposition plots were selected for inclusion
                                    as illustrations from case study computations (Figures 1 to 6). These
                                    plots contain isopleths of the logarithm (base 10) of the  cumula-
                                    tive deposition expressed in g/m2 resulting from a 1 g/sec release
                                    of material from the incinerator ship. Deposition properties were

                                                                       INCINERATION    467
11/12/85
11/12/85
11/12/85
11/12/85
11/12/85
11/12/85
11/12/85
11/12/85
11/12/85
11/12/85
11/12/85
11/12/85
11/12/85
11/12/85
11/12/85
11/12/85
11/12/85
11/12/85
11/12/85
11/12/85
500
600
700
800
900
1000
1100
1200
1300
1400
1500
1600
1700
1800
1900
2000
2100
2200
2300
2400
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
7
7
9
9
9
8
6
5
5
4
2
1
1
2
3
3
3
4
6
7
90
70
50
60
60
60
60
60
60
90
90
120
150
190
220
190
190
190
200
200

-------
selected to characterize a fine-particulate or slowly depositing gas
[i.e., most chemical forms of polychlorinated byphenyl (PCB)].
These plots cover a 100- by 100, km area with the incinerator ship
in the middle of the area.
   For the first case, the emissions from a 1-hr burn were modeled.
The cumulative buildup of plume constituents at the sea surface
for a 1-hr burn starting at 1500 hrs is shown in Figures 1, 2 and
3. In the second case, the emissions were modeled from a burn
starting at 0500 hr and continuing though the modeling period.
Different ambient meteorological conditions were selected  for
modeling these two cases.
    4)
    u
   .c
    O
   z
                                       Deposition at
                                       Hour 1600
                                       (Release Times
                                       1500 to  1600)
                                  Release Point
                                      Distance, km
                                       I.I.I
                                       0   10   20
                      East-West Distance

                          Figure I
         Cumulative Ocean Deposition 2 Hours after Start
                Of 1-Hour Incinerator Operation
8
•*

5
ฃ
O

r
o
                                       Deposition at
                                       Hour 2100
                                       (Release Time
                                       1500 to 1600}
                                                                           I   I  I   1  1   I  7  I  I
                                                                                                    I  I   I  I   I  I
                                       Distance, km
                                        t  i   i  i  i
                                        0    10  20
                                   i  i  i  i  i  i i  i
                       East-West Distance

                          Figure 3
         Cumulative Ocean Deposition 6 Hours after Start
                Of I-Hour Incinerator  Operation
  The first case shows the ability of MESOSEA to account for
temporal changes in wind speed and direction. Although the plume
is carried off to the west of the ship during the time (hour 1500
in Table 3) of the one-bum period (Figure 1), the majority of depo-
sition from this plume occurs north of the ship as a result of a
shift in winds during subsequent hours (Figures 2 and 3).
  The second case (Figures  4. 5 and 6) shows the cumulative
buildup  under higher wind conditions than the first case. A shift
in wind  direction occurs between Figures 4 and 5 and results in
a shift in the location of the deposition pattern. The steady wind
between Figures  5 and  6  results in a cumulative overlap of the
deposition.
    u
    CO
    O
   V)
                                     I  T T  T  I   I  I  I
                                       Deposition at
                                       Hour 1900
                                       (Release Time
                                       1500 to  1600)
                                              i   i  i  i
                                       Distance, km
                                        i  i  i  i  i
                                        0    10  20
          I   I  I   I  I  I  I  I   i
                                   I	i
                                          I  I  i   i  I  i
                      East-West Distance

                           Figure 2
          Cumulative Ocean Deposition 4 Hours after Start
                 Of 1-Hour Incinerator Operation
   o>
   u
   to
  O
  ฃ

   o
  (/)
  i:
   b.
   o
                                       Deposition at
                                       Hour 0600
                                       (Release Time
                                       0500 to 0600)
                      East-West Distance

                          Figure 4
          Cumulative Ocean Deposition 1 Hour after Start
              Of Continuous Incinerator Operation
468    INCINERATION

-------
   u
   c
   CO
   Q

   4-1
   O


   r
   o
                                       Deposition at
                                       Hour 1000
                                       (Release Time
                                       0500 to 1000)
                       East-West Distance
                          Figure 5
         Cumulative Ocean Deposition 3 Hours after Start
              Of Continuous Incinerator Operation
   0)
   o
   as
   to
   3
   •O
                                        Deposition at
                                        Hour 0800
                                        (Release Time
                                        0500 to 0800)
                                      Distance, km
                                       1.1.1
                                           10
                                          I
                                                20

                                               '  '
                      East-West Distance

                          Figure 6
         Cumulative Ocean Deposition 5 Hours after Start
              Of Continuous Incinerator Operation
  The isopleths in these figures may be converted to the surface
flux of a specific material by factoring the emission rates. The
formula is
                                                        (3)
where:
  S = cumulative surface flux of material,

  V = cumulative surface flux from figure based on
       unit emission, g/m2

  e = emission rate of material of interest, g/s

  E = unit emission rate  = 1 g/s
  As an example, assume that 0.24 g/s of PCB is emitted (based
on a PCB mass fraction of 0.35, destruction efficiency of 0.99990
and total incinerator feed rate of 6952 g/sec). The highest isopleth
value of any of the examples is -9, which corresponds to a value
of 1.0 x  70~9g/m2 for the unit emission cumulative surface flux,
V. Substituting into the equation gives a maximum PCB cumula-
tive surface flux of 2.4  X 10~10g/m22 in Figures 5 and 6.
  In addition to the cases shown, the first case was also run for
unstable  (Class  C) and stable (Class E) atmospheric conditions.
Both cases resulted in greater deposition than shown for Class D
conditions. Under stable conditions, there was less plume rise,
which gave higher near-surface concentrations and greater depo-
sition. Under unstable conditions, the plume mixed down to the
ocean surface faster, resulting in greater deposition rates closer to
the ship.

DISCUSSION
  The  MESOSEA model represents an intermediate  modeling
approach between screening models that provide maximum rates
and complex models  that attempt to emulate detailed aspects of
three-dimensional flow  fields  in  the  marine  boundary  layer.
Although the proposed incineration sites are remote, sufficient in-
formation generally can be obtained to run the MESOSEA model
on a case-study basis. The main parameters of surface wind speed,
wind direction, air temperature and water temperature are availa-
ble from NOAA for various at-sea locations and time periods.
However, the mixing height and precipitation data generally are
available only from  distant land, based  stations  and must be
estimated from  climatological data for the at-sea sites.
  In any application of models based on average plume charac-
teristics,  short-term departures of predicted and actual plume
properties can be expected. MESOSEA should only be used to
predict average  plume position and dispersion over time periods
of 1 hr or greater. Gaussian model testing has shown that the longer
the averaging period, the better the agreement of computed and
measured values.
  Another  limitation of the Gaussian simplification of the flow
field is the possible existence of nonrandom dispersion processes.
The model is not expected to perform well under conditions with
nonrandom flow fields. For example, the model should not be
applied to releases within a  complex shoreline atmospheric flow
pattern; all the proposed incineration sites are far enough from
shore that effects from shoreline flow patterns are not expected.
  However, nonrandom processes over the open ocean can occur
under certain  conditions. If non-random processes such as roll-
vortices are occurring, they will mainly affect the plume charac-
teristics before the plume is  uniformly mixed in the atmospheric
boundary layer (the  layer between the surface and a elevated
inversion layer).  Once the plume has reached its greatest vertical
extent, nonrandom vertical circulations will not be as important.
  The main effects of nonrandom processes will be enhanced plume
dispersion rates and  some uncertainty  in the predictions of the
plume's vertical center. The former effect will result in a tendency
for over-prediction of airborne concentrations and surface depo-
sition values.  The latter effect, although suggesting an under -
prediction in the variability in predicted over-water concentrations,
is not expected to greatly change the predicted average values at
distances beyond where an initial plume touchdown on the surface
is predicted. Hence, MESOSEA should provide reasonable esti-
mates that may be slightly conservative under conditions with non-
random air circulations. However, model outputs showing very
distant or no plume surface contacts should be regarded with some
suspicion. Nonrandom circulations not accounted for in the model
could carry the plume to the surface. This limitation applies to both
land- and water-based Gaussian models.
  The current version of MESOSEA is formulated to apply only
to over-water surfaces Applications with significant overland trajec-
tories are inappropriate.
  Keeping in  mind these limitations, the MESOSEA model can
                                                                                                     INCINERATION    469

-------
be a useful tool for studying potential plume behavior. Intensive
case studies can be made of air concentrations and deposition rates
resulting from incineration-at-sea activities.

CONCLUSION
  Although  not  currently  a designated  regulatory  model, the
MESOSEA model appears to be a potentially useful tool in studying
the  plume  potential  characteristics during  incineration-at-sea
operations.  The implementation of MESOSEA on an IBM-PC
makes the model portable. Simulation runs with this model require
only a few minutes of computer time for each modeled hour. Run-
time options allow representation of concentration fields as both
real-time plots and/or data files. The real-time plots allow real-
time interactive applications of MESOSEA. The data file output
provides both an archive of model outputs as well  as data for
making detailed plots.

ACKNOWLEDGMENTS
  The authors wish to acknowledge J. V. Ramsdell for providing
a  personal  computer  version  of MESOI and giving  helpful
comments that greatly expedited the development of MESOSEA.
This work was supported by the U.S. EPA under contract num-
ber  68-00-3319.

REFERENCES
1.  Congress of the United States, Office of Technology Assessment, "Ocean
   Incineration: Its Role in Managing Hazardous Waste," OTA-0-313,
   Washington, D.C. 1986.
2.  Ramsdell, J. V., Athey, G. F. and Glantz C. S "MESOI Version 2.0:
   An Interactive Mesoscale Lagrangian Puff Dispersion Model With Depo-
   sition and Decay," prepared by Pacific Northwest Laboratory for the
   U.S. Nuclear   Regulatory  Commission, Washington  D.C.,
   NUREG/CR-3344 (PNL-4753) 1983.
3.  Petersen, W. B.  and Lavdas, L. G. "INPUFF 2.0—A Multiple Source
   Gaussian Puff Dispersion Algorithm User's Guide", U.S. EPA Research
   Triangle Park. NC. EPA-600/8-86-024,  1986.
4.  Petersen W. B., "User's Guide for PAL, Gaussian-Plume Algorithm
   for Point, Area, and Line Source," U.S. EPA Research Triangle Park,
   NC, EPA-600/4-78-013. 1978.
5.  Bowers, J. F., Bjorklund, J. R. and Cheney, C. S., "Industrial Source
   Complex (ISC) Dispersion Model User's Guide." U.S. EPA Research
   Triangle Park, NC. EPA-450/2-77-018 1979.
6.  Droppo, J. G. Vail, L. W. and Ecker, R. M.,  "INSEA User's Manual
   Environmental Performance Model of Incineration-at-Sea Operations."
   U.S. EPA, Washington, D.C., draft report,  1987.
7.  Start, G. E. and Wendell, L. L, "Regional Effluent Dispersion Calcu-
   lations Considering Spatial  and  Temporal Meteorological Variations
   NOAA  Technical Memorandum," ERL ARL-44, 1974, 63p.
8.  Pasquill, F., "Atmospheric Dispersion of Pollution." Quart. J. Roy.
   Meteor. Soc., 97:1961 369-395.
9.  Hasse, L. and Weber H. "On the Conversion of Pasquill Categories
   for Use over Sea." Bound.-Layer Meteor. 31:1985 177-185.
10. Joffre, S.  M., "The Structure of the Marine Atmospheric Boundary
   Layer a Review from the Point of View of Diffusivity, Transport and
   Deposition Processes,"  Finnish Meteorological Institute, Technical
   Report No. 29.  Helsinki  1985.
11. Garratt, J. R, "Review of Drag Coefficients over Oceans and Con-
   tinents,' Mo/i.  Wea.  Rev. 105.  1977. 915-929.
12. Businger. J. A., Wyngaard, J. C.. Izumi Y. and Bradley. E. F., "Flux
   Profile Relationships in the Atmospheric Surface Layer." J. Atmos.,
   Sci., 28:1971 181.
13. Paulson.  C. A., "The Mathematical Representation of Wind Speed
   and Temperature Profiles in the Unstable Atmospheric Surface Layer."
   J. Appl. Meteor. 9:1970 857.
14. Sehmel G. A. and Hodgson. W. J.. "A Model  for Predicting Dry Depo-
   sition of Particles and Gases to Environmental Surfaces," Pacific North-
   west Laboratory. WA. PNL-SA-6721. 1978.
15. Van Voris. P.. Page, T. L., Rickard,  W. H., Droppo, ;. G. and
   Vaughan B. E., "Environmental Implications of Trace Element Releases
   from Canadian Goal-Fired Generating Stations, Phase II, Final Report,"
   Volume II, Appendix B, Canadian Electric  Association, Montreal,
   Quebec. 1984.
470     INCINERATION

-------
             Risk  Analysis  of Pollutants  at Hazardous  Waste  Sites:
                             Integration  Across  Media is the Key
                                             Lyse D. Helsing, Dr. P.H.
                                             Mary P. Morningstar, Esq.
                                             Joan B. Berkowitz, Ph.D.
                                             Risk  Science International
                                                  Washington, D.C.
                                               Thomas T.  Shen, Ph.D.
                                                 Columbia University
                                                New York, New York
ABSTRACT
  Most pollutants at hazardous waste sites are multimedia pollu-
tants. Risk analysis can be used as a technique to fully address risks
associated with hazardous waste in all environmental exposure
pathways; i.e., for integrating across media and pollutants. The
transfer and transformation potential of many pollutants from one
medium to another must be taken into consideration, however,
for environmental contamination to be fully assessed and managed.
  The paper describes types of pollutants most frequently found
at hazardous waste sites and shows how they contaminate all media
(water, air and soil) and may cause adverse health effects to workers
and surrounding populations through all exposure pathways (in-
gestion, inhalation and dermal contact). Techniques currently used
to identify and quantify pollutants and new procedures developed
by U.S. EPA to better account for pollutant transport through
various media are discussed briefly, as are environmental statutes
and the implementing U.S.  EPA  rules that regulate hazardous
wastes.

INTRODUCTION
  Pollution from hazardous waste is one of today's major environ-
mental problems. An important and often overlooked part of that
problem is the fact that pollutants cross from one environmental
medium to another, causing damage to the air, water, soil and biota
through which the pollutants may cycle. Hazardous waste must
be managed as a multimedia problem,  not solely as a water
problem, air problem or  soil problem. While it is  widely known
that pollutants are transferred from one medium to another and
may change form, knowledge of  how this happens and how it
adversely affects health and the environment and those who live
in it is  not widely reflected in the regulatory process. In fact, to
date federal and state water, air and hazardous waste laws primarily
have addressed cleanup of pollutants in a specific medium, losing
the vital perspective of the environment as an integrated ecologi-
cal whole. As a result, pollution control methods frequently have
become in effect pollution transfer methods that have increased
rather than reduced exposure to a pollutant.

TYPES AND FATE OF POLLUTANTS AT
HAZARDOUS WASTE  SITES
  The pollutants that are most frequently detected in mixtures of
waste materials deposited  at chemical disposal sites are multimedia
pollutants. Many of these have been designated as "hazardous
wastes" by the U.S. EPA under RCRA. Based on their toxicity,
ignitability, corrosivity or reactivity, a total of 389 chemicals are
now listed as Hazardous Constituents. One subgroup includes 113.
chemicals designated as Acute Hazardous Wastes because they are
fatal to humans in low  doses or because they are  "capable of
causing or significantly contributing to an increase in serious irre-
versible or incapacitating reversible, illness." Another subgroup
of this list includes 239 chemicals listed as Toxic Wastes shown
in scientific studies to have "toxic, carcinogenic, mutagenic or tera-
togenic effects on humans or other life forms. "(1)
  Several lists have been developed to tabulate chemicals found
at or in the vicinity of chemical disposal sites. Many of the lists
pertain to a single waste disposal site; others pertain to waste dis-
posal  sites within individual states. The Mitre list assembled for
the U.S. EPA in 1983(2) is more extensive in terms of geographic
distribution of sites and includes more types of industrial facili-
ties than would be present in any one state. It lists 230 chemicals
found near  546 NPL sites. According to this list, 173 chemicals
were reported in groundwater in the vicinity of chemical waste dis-
posal  sites, 162 in surface water and 65 in air. A 1985 updated list
includes 465 different substances identified at 818 NPL sites.
  Table 1 lists the 38 chemical substances most frequently found
at hazardous waste sites. The list is broken into five principal
groupings:

  Halogenated aliphatic compounds
  Halogenated aromatic compounds
  Non-halogenated aromatic compounds
  Miscellaneous organic compounds
  Trace metals and other inorganic compounds

  It is based on information provided in the various lists and is
adapted from  a list assembled  by the Executive Panel on the
"Health Aspects of the Disposal  of Waste Chemicals."3 These 38
chemicals were selected primarily on the basis of their reported
presence at numerous chemical disposal sites, but also because they
have a potential for transport through the environment and for
causing adverse health effects. As  illustrated in Table 1, most chem-
icals listed have in fact been found in groundwater, surface water
and air at or near hazardous waste sites. While data on the presence
of these chemicals in the soil have not been as systematically tabu-
lated, available documentation on their fate and transport would
seem  to indicate that most also  may be present in the soil at or
near hazardous waste sites and therefore have been used as the basis
for inclusion in the table.4
  Table 2 lists health hazards associated with the same 38 chemi-
cals and shows that 18 compounds are carcinogenic; i.e., either
a "known or suspected human carcinogen or shown to be carcino-
genic  in more than one species or sex in an animal bioassay or
shown to increase the incidence of site-specific malignant tumors
                                                                                                  MULTI-MEDIA     471

-------
                            Table 1
       Chemicals Frequently Found at Chemical Disposal Sites
Coopounda
HalotflnaUd Utphatlc Coapourrfa
    Methylene Chloride
    Chloroform
    Carbon tetraehlorlde
    1,1-Dlohloro* thane
    1,2-Dlchloro* thine
    1,1, 1-TrlcMoroethane
    1 1 1 ,2-Trlohloroethane
    1.1.2 ,2-Tatraซhloroethane
    Vinyl chloride
    1.1-Dlohloroathylana
    trana-1 ,2-Dlchloroซthylene
    Trlotiloroethylene
    Tatraoh loroethj ItfM
                            GroundMater
    Chlorobenaana
    1 .2-Dlchlorobenzene
    P an tach loro phenol
    2,3,7,8-Tetraohlorodlberao~p-dloxin
    Polyohlor Inated blphenyla
    DOT
            Aroemtlo Coapowida
    Benzene
    Toluene
    Xylcne
    Ethlybeniene
    Phenol
    Naphthalene
    Benzo{a]pyrene

 HlacclLancouj Qrfftnlo Coapounda
    AM tone
    Methyl ethyl ketone

 Trace HeUia and Other Inorjsnlc
    Arsenic
     Copper
     Cyanide (frซe cyanide ion)
     Lead
     Hlckel
     Zinc
                                       Surfftotuetar
Soil

  X
in a single species or sex and for which there is a statistically sig-
nificant  dose-response relationship in more than  one exposed
group. "(5) Five of these 18 chemicals are proven carcinogens in
humans: vinyl chloride, benzene, arsenic, cadmium and asbestos.
Table 2 also shows that 18 of the chemicals most frequently found
at or near hazardous waste sites are classified as teratogens and
reproductive toxins; 13 are classified as mutagenic; 2 are consi-
dered to be acutely toxic; and 13 are considered to cause chronic
toxicity.
                             Tabk 2
        Chemical Frequently Found al Chemical Disposal Sites
                                       lolly
                                                       Aeult     Qirvtle
                                                       Tป telly   Iff.ct
    .
   1 , 1 , 1-T r tcMerovtmn*
   1 . 1 ,2-TM*Ottnซ
   Vinyl oMorM*
   l,l-0lehloroซuปyl*>ซ
   trwu-l ,2-DtchlorMthylcnc
   TrlctilorwtliylMW
   Tซtr*<* lorovttqr 1*M
   PcnUeMorophwwl
   i.3,?,*-T*lrซohliปr
   Poj)rchlorlftซtซd bl
   DDT
   Bcnient
   Toluan*
   t*nioUlpyrซn<

 NlMvlUrwoii Or^nlo Camoountt

   Methyl ethyl ice ton*
    IMrour,
    Mick* I
    Zinc
TRANSFER AND TRANSFORMATION OF POLLUTANTS
AND POTENTIAL ADVERSE EFFECTS

Transfer of Pollutants
  Pollutants do not remain where they are initially deposited, but
rather move from the source to receptors by many routes including
air,  water and land, according to management practices used.
Transfer of pollutants to another medium may result from such
pollution abatement  devices as air scrubbers which increase the
volume of waste substantially; e.g. from 3 to 6 tons of scrubber
sludge may be produced for every ton of sulfur dioxide removed
from flue gas.*
  During treatment of hazardous wastewater, water pollutants may
be volatilized into the air or collected and disposed of on land as
sludge. In turn, the sludge contaminates the land and the ground-
water through residues. In cases where sludge is disposed of through
incineration,  pollutants may be transferred to the air through
emissions.
  A study of 27 of the approximately 200 chemicals found at Love
Canal shows that 18 of the 27 chemicals were identified in air, water
and soil.7 Contamination of ground water by leaching from land-
Tills or surface impoundments is another important source of pol-
lutant transfer.
  Pollutants  may be transferred to another medium by incinerating
wastes. As discussed elsewhere,1 pollutants of most concern from
incineration   are particulars, heavy metals  and acid gases.
Depending on the waste burned and the process used, other pollu-
tants also may be emitted. Metals in the waste will not be destroyed
by incineration and will leave  the incinerator as gases,  fly ash or
other residues that may cycle  into the environment and settle in
water or on  land.
  Hazardous waste also may volatilize and be transferred to other
media.  Volatilization from waste piles, waste treatment processes,
lagoons and landfills can be an important source of pollutants
across different media at chemical disposal sites. Volatilization is
a process by which a substance is transferred from a liquid or solid
phase to a vapor phase; it occurs readily from most organic com-
pounds at a  relatively low temperature. Studies have shown that
even very stable chlorinated hydrocarbons with very low vapor pres-
sure, such as PCBs and pesticides, do volatilize.(9) In fact, field
monitoring data have shown that PCB concentrations were fairly
high in ambient air and vegetation near PCBs dump sites and cer-
tain contaminated dredge spoil sites along the Upper Hudson River
of New York State.
  Vaporized contaminants of particular environmental concern at
chemical disposal sites are the halogenated organics and the aro-
matic hydrocarbons. According to the 198S U.S.  EPA survey of
substances found at  proposed and  final NPL sites, the  most
ubiquitous substances include trichloroethylene,  lead, toluene, ben-
zene, chloroform and polychlorinated  biphenyls (PCBs).  Tri-
chloroethylene (TCE), for example, rapidly volatilizes into the
atmosphere,  where it reacts with hydroxyl radicals to produce
hydrochloric acid, carbon monoxide, carbon dioxide and carboxylic
acid. This vaporization/transformation is thought to be the most
important transport and fate process for TCE in surface water and
in the upper layer of soil. TCE adsorbs to organic materials and
can be accumulated.  TCE also leaches into the ground water fairly
readily and  is a common contaminant of groundwater around
hazardous waste sites.(11) TCE is carcinogenic to mice via oral
administration, producing hepatocellular carcinomas. It also has
been found to be mutagenic using several microbial assay systems.
Following chronic exposure to levels greater than 2,000mg/m3,
TCE has been shown to cause renal toxicity, hepatotoxicity, neuro-
toxicity and dermatological reactions in animals.12
   A study of chemical and physical processes of 43 chemicals
indicated  that  for  volatile  chemicals,  such as acrynolitrile,
 1,1-dichloroethylene, ethylene chloride,  perchloroethylene and
benzo(a) pyrene, chemical removal residence times range between
3  and  70 days, based on reaction with hydroxyl radicals and
472    MULTI-MEDIA,

-------
ozone.(13) These data show that vaporized contaminants may travel
from rural to metropolitan areas, thus causing pollution problems
not only at the source but also in distant regions. Moreover, the
hazardous properties of most organic compounds probably will
persist unless destroyed by reactions. The undestroyed hazardous
vapors  and gases may  be adsorbed  on  small  particles  in the
atmosphere and then will fall out on land and surface waters. Sub-
sequently, through a cyclic process, they can be readmitted into
the atmosphere.

Transformation of Pollutants
  Other substances frequently found at hazardous waste sites that
also pose serious environmental and health concerns include trace
metals and other inorganic compounds such as arsenic, cadmium,
chromium, cyanide, mercury, nickel and asbestos. As illustrated
in Table 1, all of these compounds have been found in all media
at or near hazardous waste sites.  Table 2 illustrates the  health
hazards from these various compounds.
  Determination of accurate exposure from different media is com-
plicated for many pollutants due to chemical speciation. In fact,
pollutants such as lead, arsenic, mercury and cadmium may appear
in different species both in exposure media and in the organism,
thus complicating the multimedia risk assessment. For example,
in the natural environment, arsenic has four different oxidation
states (-3, 0,  +3 and +5) and chemical speciation is important
in determining its distribution and mobility. Interconversions of
the + 3 and + 5 states, as well as organic complexion, are the most
important. Arsenic is generally quite  mobile in the environment
because it is  metabolized to organic  arsenicals by a number of
organisms, hi the aquatic environment, for instance, volatilization
is important when biological activity or highly reducing conditions
produce arsine or methylarsenics. Sorption by the sediment is an
important fate for the chemical. Its ultimate fate is probably the
deep ocean, but it may pass through numerous stages before finally
reaching the  sea floor.14
   Determination of total arsenic in air, food and water for health
risk assessment gives limited information since arsenic may appear
in many  species. As illustrated in Table  3, in  air  and  water,
inorganic arsenic in the trivalent or pentavalent state is dominating;
in food from terrestrial sources, methylated forms are the major
sources; in seafood, there are  other organic forms.

                            Table 3
    Prevalence of Forms of Arsenic and  Mercury in  Various Media
                   Air
                           Water
                                     Soil, Plants    Fish  Hunan Urine

                                                    (*)

                                                    (*>
As+5

Methyl As

Arsenobetaine

Hgฐ



MeHg
ป*, +, and (*) Indicate relative concentrations in exposure media and
urine.
Source:    Plscator, M., "Factors of Importance for Planning and
         Interpretation of Population Surveys, "Multimedia
         Approaches to Assessment and Management of Hazardous
         Air Contaminants, A Report from NATO/CCMS Workshop,
         Denver, CO, Oct. 8-10, 1986, p. 17.


  There  are large differences in toxicity between the different
species of arsenic: trivalent arsenic  is highly toxic; however,
arsenobetaine,  a species occurring in seafood, is nontoxic to
humans.  Certain foodstuffs also may contain organic arsenicals
used as growth promoters for animals such as chickens, hi humans,
arsenic has a complex metabolism. Inorganic arsenic is methylated
and the major part of the arsenic excreted in urine is in this organic
form. Arsenobetaine is easily absorbed from the gut, but it is not
metabolized and is excreted unchanged in urine. Thus,  determi-
nation of total arsenic in urine could be misleading: determina-
tion of species will give .more information for assessment of
exposure and risk.15
  Another good example of chemical speciation and  the difficulty
posed for a multimedia risk assessment  is mercury, which may
occur in relatively low concentrations in ambient air  and does not
cause systemic effects. Increases in basic foodstuffs from long-term
deposition to soil or water, however, must be taken  into account
to determine mercury accumulation in the body. A further exam-
ple of a multimedia conversion process is methylmercury, which
is formed by microorganisms in water from inorganic  mercury
derived from mercury vapor in air,  ultimately leading to high con-
centrations of methylmercury in fish. Methylmercury is not excreted
in urine and must  be measured in  human blood or hair. In fact,
hair has been found to be an excellent indicator medium for esti-
mates of the body burden of methyl mercury.16


MULTIMEDIA RISK  ANALYSIS
  Risk analysis can be an excellent mechanism for integrating
across media and pollutants. To date, however, techniques used
to assess environmental and health risks have been geared to iden-
tification and quantification of pollutants in a single medium—
water, air or soil—without taking into full  consideration the trans-
fer  of pollutants from one medium to another and the chemical
speciation of various compounds in different media. Ideally, then,
risk analysis should take into account multiple emission sources;
estimate contaminant concentration in all media resulting from
emissions  and  discharges; define  multiple routes of  exposure;
estimate exposure to humans, animals and plants; and come up
with control methods that reduce exposure to pollutants.
  Human and non-human exposure assessment is the most criti-
cal  step of a multimedia assessment and needs to include source
assessment, pathways of fate analysis, estimations of environmental
concentration, population analysis  and integrated exposure anal-
ysis. Multimedia exposure assessment is very difficult to achieve,
however,  since estimation  of a pollutant  concentration at an
exposure point must rely on field data and  mathematical modeling
results.
  Because field  data on all parameters needed to identify and quan-
tify exposure frequently are not available,  heavy reliance has been
placed on mathematical models. There is, however, no ideal mathe-
matical model that can identify and quantify all needed parameters,
and multiple model usage is  necessary to account for pollutant
transport through  different media.
  To assure consideration of the most appropriate mathematical
models for each exposure scenario (surface  water, groundwater and
air), the U.S. EPA recently developed and published  guidelines
for mathematical model selection. As discussed elsewhere,17 these
guidelines are incorporated  into an overall process by  which  a
problem is identified and a model is  selected. They include problem
characterization, site characterization, model selection criteria, code
installation and model application.
  Government  agencies and lawmakers must recognize the need
to use a multimedia assessment approach for characterization of
environmental  and health risks from exposure to pollutants at
hazardous waste sites. This first essential  step can lead to the use
of management approaches that also take into account the environ-
ment as an integrated ecological whole. Integration of a multimedia
perspective in the risk assessment and management  process also
will stimulate development of more targeted techniques and field
and health data banks. Control practices that address the multi-
media aspect of environmental contamination also need to be
implemented by the U.S. EPA as regulatory guidelines for full in-
tegration at hazardous waste sites.
                                                                                                        MULTI-MEDIA    473

-------
REGULATORY CONTROL ACTIONS NEEDED
  Since CERCLA was enacted in 1980, the U.S. EPA has spent
millions of dollars investigating options to most efficiently clean
up contaminated properties. The principal obstacle to effective
remediation  at  these sites has been the lack of any cohesive
approach to address pollution affecting all media — air, water,
soil and biota. Nevertheless, the U.S. EPA has discovered recently
that the management of risks at actual Superfund sites may neces-
sarily lead to a  multimedia approach.
  The U.S.  EPA is attempting to remediate contamination at
hazardous sites by developing techniques to compare risks in differ-
ent media to develop the most efficient set of control options likely
to achieve a degree of risk reduction." U.S. EPA officials also
have been listening to their contractors. One U.S.  EPA contrac-
tor developed a multimedia scheme to rank the relative  risks
associated with a Superfund site.(19) The approach is implemented
on a case-by, case basis.  Moreover, the only effective implemen-
tation has been retroactive—cleaning up already contaminated
hazardous properties. Piecemeal legislation and specificity of re-
quirements inhibit the integrated analysis, control and enforcement
of the multimedia approach.
  Congress amended CERCLA in October 1986 by passing SARA.
Congress recognized the  U.S. EPA's attempt to control environ-
mental pollution across all media, so SARA requires that the U.S.
EPA elect permanent remedies using alternative treatment or
resource recovery technologies as much as possible.20 The removal
action provision mandates:

    .. .any removal action undertaken by the President under
   this subsection should contribute to the efficient performance
   of  any long-term action with respect to the release or
   threatened release concerned.21

In addition.  Congress mandated that the U.S. EPA select cleanup
actions that "permanently and significantly reduce the volume,
toxicity and mobility of the hazardous substances" at a Superfund
site.22
  In some respects, SARA codified the U.S. EPA's policy in rela-
tion to compliance with other environmental laws during a cleanup
of a hazardous  site. For example,  Section 121 requires the U.S.
EPA to determine whether water quality criteria under the Clean
Water Act (CWA) are "relevant and appropriate under  the cir-
cumstances of a release or a threatened release."23 Contaminat-
ed material must be removed to a facility that is in compliance with
RCRA.24 Finally, the U.S. EPA's groundwater protection strategy
for cleanup  under SARA relies on the groundwater protection
standards developed under RCRA which, in turn, were developed
using the maximum contaminant levels (MCLs) that the U.S. EPA
promulgated pursuant to  the Safe Drinking Water Act (SOWA).25
  Notwithstanding this apparent overlap of activity, it still appears
that pollution control and risk abatement are studied on a single
medium basis. Superfund's implementation of the U.S. EPA and
state  agencies has attempted  to  remediate contamination and
mitigate risk  by recognizing retroactively that a pollutant can affect
all media. The other major environmental statutes do not address
pollution across all media as cohesively as SARA.
  RCRA, passed in 1976 and amended by the Hazardous and Solid
Waste Amendments (HSWA) in 1984, mandates the "cradle-to-
grave" management of hazardous waste from its generation to its
disposal. The U.S. EPA  has promulgated a vast body of regula-
tions to implement that Congressional mandate. These regulations
are designed primarily to protect  two media—the soil and the
groundwater. More recently, however, the U.S. EPA proposed
regulations under the Clean Air Act (CAA) that would apply to
emissions  from  RCRA facilities.
  Hazardous waste landfills must go through a lengthy permitting
process to ensure that the waste they accept will not adversely
impact those media. RCRA thus appears to focus on minimizing
risks from wastes,  but no RCRA regulations address the risks
associated with the volatization of the landfilled hazardous wastes
and their subsequent impact on the air.26
  There are two provisions in RCRA that implicitly address multi-
media assessments.  RCRA Section  7003—the Endangerment
Assessment Provision — mandates that the U.S. EPA determine
the magnitude and probability of actual or potential harm to health
or the environment by a threatened or actual release of hazardous
waste.27 In general, an cndangerment assessment must identify:2*
  Hazardous wastes  present in all media
  Environmental fate and transport mechanisms within the media
  Intrinsic lexicological properties of human health standards
  Exposure pathways and  the extent  of  expected  or potential
  exposure
  Populations at risk
  Extent of expected harm
Since  the danger must already exist before an assessment is per-
formed, this RCRA provision is, like Superfund cleanups, retro-
active; it does not address  the importance of prospective risk
management to protect all media.
  The Exposure Information/Health Assessments section in the
HSWA appears to address multimedia pollution. Section 3019(a)
provides that each application for a landfill or surface impound-
ment permit filed after July  1985 must be accompanied by infor-
mation on the "potential for the public to be exposed to hazardous
waste  or hazardous constituents through releases  from the unit."29
An applicant  must specify  the potential  pathways of human
exposure and the nature of that exposure.30 This provision seems
to recognize that few pollutants stay in one medium and that they
can be transferred to another or be transformed into other sub-
stances that may have different or  more deleterious effects.
  It is only in this section that Congress attempted to address the
multimedia approach. Since RCRA's focus is to minimize risks
from  hazardous wastes, it may be prudent for the U.S. EPA, in
lieu of a statutory change, to implement a multimedia approach
in its  permitting of hazardous waste generators and treatment,
storage and disposal facilities. The U.S. EPA has already begun
such an approach. Since the  agency became aware of potential air
emission problems from landfills, it has conducted studies to evalu-
ate the extent of the problem and the circumstances that could cause
harm  to human health and  the environment.31
  Section  104 of the CWA provides  for  the establishment of
national programs to prevent, reduce and eliminate pollution in
cooperation with other federal,  state and local agencies.32 One
section of the CWA seems  to  encourage the multimedia perspective.
Section 40S(d) requires discharge permits for municipal treatment
plants to incorporate conditions on the management of sludge and
provides guidelines for that management.33 Nevertheless, the prin-
cipal focus of the CWA, as well as the U.S. EPA's implementing
regulations, is the prevention of discharges into surface water. The
law seeks to control pollutants as if they remain  in the same
medium—water. Prohibiting direct discharges into a lake may not
be enough to protect the lake from pollution.
  Congress could amend the CWA to specifically address the mul-
timedia approach by  looking to the states' perspectives on multi-
media. Some states have  begun to adopt regulatory approaches
which specifically deal with intermedia impacts or trade-offs. The
governors of the Great Lakes states adopted TSCA which directs
the states  to work together  toward the development of compre-
hensive, integrated and complementary policies on the control of
toxic  substances.34 The Agreement's emphasis on pollution con-
trol through waste reduction illustrates that controlling the dis-
charge of pollutants does not sufficiently reduce the risks posed
by those pollutants.
  The TSCA attempts to induce coordination among pollution con-
trols for different media. Section 9(b)  mandates the U.S. EPA
Administrator to coordinate actions taken under  federal laws
administered by the U.S. EPA.35 If the U.S. EPA determines that
a risk  to health or the environment associated with a chemical sub-
474     MULTI-MEDIA

-------
stance or mixture could be eliminated or reduced to a sufficient
extent by actions taken under the authority of a different law, the
U.S. EPA can use that authority to protect against those risks.
This provision gives the U.S. EPA the necessary authority to
develop a multimedia approach to control risks from pollutants.
TSCA could be strengthened and viewed as the law of first choice
to control pollutants, instead of as the last resort, as has frequent-
ly happened in recent years.
  TSCA seems to be the only statute that already has the statu-
tory language that the U.S. EPA could use to develop organic regu-
lations to protect the soil, air and water. It is possible that there
would be no need to amend other statutes; the language of Section
9(b) could provide the U.S. EPA with the framework for a national
pollution control system, instead of reinforcing its current feeling,
based on other statutes, that separate controls are sufficient to pre-
vent the pollution of our air, water  and soil. The U.S. EPA  can
use its existing pollution control practices under RCRA, CWA,
TSCA, CERCLA and CAA, to take into account the interrela-
tionship of air, water and soil pollution. By using the authority
in Section 9(b) of TSCA, the U.S. EPA could then implement those
control practices as regulatory guidelines to effectively control
multimedia pollutants. The U.S. EPA,  however, is unlikely to act
independently. Congress may have to provide the U.S. EPA with
a direct mandate  to pursue an aggressive program to control mul-
timedia hazardous  waste pollution.
REFERENCES
 1. Code of Federal Regulations, 40 CFR Part 162.
 2. Mitre Corporation. Computer printouts of National Priori-
    ties List data summaries for 546 final and proposed sites and
    881 sites currently on data base as of Sept. 7, 1983. Prepared
    for the U.S. EPA, Washington, DC.
 3. Universities Associated for Research and Education in Pathol-
    ogy, Inc.  Draft Report, Oct. 1984, 2-15.
 4. U.S. EPA, Chemical, Physical & Biological Properties of Com-
    pounds Present at Hazardous Waste Sites, Report No. 9850.3,
    prepared by Clement  Associates, May 30, 1986.
 5. Ibid., Appendix A. Supra,  fn. 3, pp. 2-3 to 2-22. J. Doull,
    C.D. Klaassen, & M.O. Amdur, eds.,  Casarett and Dpull's
    TOXICOLOGY, The Basic Science of Poisons, 2nd ed., Mac-
    millan Publishing Co., (1980)
 6. Conservation Foundation, "Controlling Cross-Media Pollu-
    tants," State of the Environment,  (1984) 326.
 7. Kim, C.S. et al., "Love Canal: Environmental Studies.in"
    Hazardous Waste Disposal: Assessing the Problem, ed.  J.H.
    Highland, Ann  Arbor Publishers,  Ann Arbor, MI, (1982)
    83-94.
 8. Helsing, L.D. and Shen, T.T. "Multimedia Approaches to Air
    Pollution Control in Hazardous Waste Treatment and Dis-
    posal," Proc. of the 1987 Specialty Conference, Environmental
    Engineering, John D. Dietz, Ed, (1987) 399-405.
 9. Shen, T.T. and Tofflemire, T.J., "Air pollution Aspects of
    Land Disposal of Toxic Waste," J. of Environ.  Eng.  Div.
    ASCE, 106, No. EEI, (1980) 211-226.
10. Tofflemire, T.J. and Shen, T.T., "Volatization of PCB from
    Sediment and Water: Experimental and Field Data," Proc.
    of the llth Mid-Atlantic Industrial Waste Conference, Univer-
    sity Park, PA, (1979) 100.
11. Supra,  fn. 4,  pp. 537-540.
12. Supra,  fn. 4,  pp. 537-540.
    Supra, fn. 5.
13. Cupit, L.T., "Fate of Toxic and Hazardous Materials in Air
    Environment," U.S. EPA Publication No. 600/S3-80-084 Dec.
    1980.
14. Supra,  fn. 4,  pp. 62-64.
15. Piscator M., "Factors of Importance for Planning and In-
    terpretation of Population Surveys," Multimedia Approaches
    to Assessment and  Management of Hazardous Air Con-
    taminants, A Report from NATO/CCMS Workshop, Denver,
    CO, Oct. 1986, (1987) 46-47.
16. Ibid., pp. 47-48.
    Supra, fn. 5, pp.  422-424.
17. Segna, J.J., "U.S. the U.S. EPA Guidelines For Mathema-
    tical Model Selection for Performing Exposure Assessments,"
    Proc.  of  the  1987  Specialty Conference, Environmental
    Engineering, John D. Dietz, Ed, pp.  714-719
18. Supra, Fr. 6,  p. 351.
19. Id., p. 351.
20. Superfund Handbook, ERT/Sidley & Austin, Concord, MA,
    (1987),  39
21. Superfund Amendments and  Reauthorization Act (SARA),
    asa!04(a)(2) (1986).
22. SARA, ง121(b).
23. SARA, ง121(d)(2)(B)(i).
24. SARA, ง121(d)(3).
25. Weishaar, M.G.  and Gray, K.F., "Groundwater Cleanup
    Under Superfund: The New Ballgame," Toxics Law Report-
    er 105 (1987).
26. U.S. EPA, Federal Register, "The Interim Status Standards
    for Hazardous Waste and Consolidated Permit Regulations,"
    (Dec. 18, 1978; May 19,  1980 and Jan. 12, 1981).
27. RCRA, ง7003, as amended (1984).
28. U.S. EPA, Office of Solid  Waste and Emergency Response
    Directive  "Endangerment Assessment Guidance," Directive
    #9850.0-1 (1985).
29. RCRA, as amended, ง3019(a).
30. Id.
31. Helsing, L. and Shen, T.T. "Secure Landfills — Hopes and
    Fears,"   Proc.  National  Conference on Environmental
    Engineering, (1983) 672
32. Federal Water Pollution  Control Act, ง104 (1978).
33. Id., ง405(d).
34. Marsh, L., "New York  State Perspectives on Inter-Media
    Approaches to Pollution  Control," Proceedings  of 1987
    Specialty  Conference—Environmental Engineering, (1987).
35. Toxic Substances Control Act,  ง9(b) (1980).
                                                                                                    MULTI-MEDIA    475

-------
                            Factors  and  Phenomena  Involved in
                               Multimedia  Exposure Assessment

                                         T. Edward Fenstermacher, Ph.D
                                                     Luca Ottinetti
                                         Pickard,  Lowe, and Garrick, Inc.
                                                   Washington, DC
ABSTRACT
  Exposure assessment in a multimedia environment requires the
understanding of the large number of phenomena which affect the
transfer of material and energy across interfaces and through
media. Different sets of physical, chemical and  environmental
factors must be considered depending on the characteristics of the
material, release and medium of concern. Before beginning an
exposure assessment, it is important to know which phenomena
and factors will be of importance  for the case at hand.
  The physical phenomena of importance in determining trans-
port of gases and liquids and paniculate solids across the inter-
faces between air, water, soil and biota are discussed, along with
the transport of materials within a given medium. In each case,
the physical, chemical and environmental factors which affect the
resulting exposure are determined.
  The importance of each factor in calculating each type of in-
termedia or intramedium transport is considered. The determina-
tion of whether a factor is of major, minor or no importance is
made on the basis  of whether a change in a given factor over its
entire  plausible range of values would result in a change of more
or less than an order of magnitude or no change at all.  The
importance of each factor is  tabulated in a matrix for each type
of intermedia or intramedium transport.

INTRODUCTION
  Since exposure assessment  is expensive and time consuming, it
clearly is important to know which of the various factors and
phenomena involved are likely to have a major impact on a given
exposure  assessment and  which  can safely be neglected.  The
purpose of this paper is to provide a basis for such decisions by
putting the number of existing phenomena, factors and models into
a common framework. A factor, as used here, is cither a quantifi-
able physical property or other information which allows the person
performing the exposure assessment to determine which model to
use. An examination of a variety of models has shown that most
models of transport and transformation in a particular medium
or across a particular interface treat the same phenomena and use
the same factors. Since it would be impossible to cite all of the
relevant literature for each phenomenon, a bibliography of useful
reference material is provided instead.
  An  exposure assessment typically mimics the structure of the
physical processes it models and thus consists of  several distinct
chronological stages. The exposure assessment stages considered
here are the source determination, the release characterization, the
evaluation of environmental transport and transformation and the
determination of the level of  human exposure. At each stage, the
types of information needed are different and the information
required depends not only on the stage, but also on such things
as the properties of the chemical released, the dominant exposure
pathways and the types of exposure which are of  concern.
  The emphasis in this paper will be on environmental transport
and transformation, but some consideration will be given to the
characterization of the source and the release, since these provide
essential input to later processes. The determination of the level
of human exposure provides a basis for the comparison of the im-
portance of various exposure pathways and will also be discussed.
Emphasis will be on the factors and phenomena which will affect
the assessment of exposure from potential releases from superfund
sites and will thus focus on processes which occur in sofl and water,
with  releases from soil  and water to air and with transfer from
soil and water to biota.

FACTORS AND PHENOMENA USED IN
EXPOSURE ASSESSMENTS
  In  the following sections, the phenomena and factors which are
important in different stages of the exposure assessment, for trans-
port and transformation in different media and for transport across
various interfaces are discussed. The results are summarized in
Tables 1 and 2. In these tables, the importance of each phenome-
non and factor discussed below in characterizing the behavior of
a chemical substance for  each  stage, media or  interface is given
a numerical  value. This value equals 2 if consideration of the
phenomenon is essential or if the impact of changes in the value
of a  factor may change the results by more than a factor of 10,
equals 1 if the phenomena should be considered in a detailed study
or if the impact of changes in the value of a factor will change
the results by less than a factor of 10, and equals 0 if the phenome-
non  or factor is irrelevant.

Factors Characterizing the Source and
Release of Chemicals
  Source characterization is the identification of the material of
concern and the determination of its physical properties. Identi-
fication is performed by determining the common name of the
material, the compounds present in the material if it is a mixture
and  the chemical and structure formulas of each compound
included. Although a seemingly trivial part of the exposure assess-
ment, identification is essential to the determination of the physi-
cal and chemical properties required in later stages.
  The source properties which will be of major significance for
assessing exposure at superfund sites include the molecular weight,
initial concentration and solubility of each component. The density
476    MULTI-MEDIA

-------
also can be important in characterizing a release in the amount
released, release rate,  release location and release temperature.
Other properties, which characterize the interaction of the chemical
with particular media  or across specific types of interfaces, will
be discussed in the appropriate sections below.

Transport and Transformation in Water
  Chemical materials  released to water, either directly or  from
another phase, consist  of solution or suspension of particles. The
movement of these materials is governed largely by the motion of
the water, which  determines both  the path  traveled by a con-
taminant and the amount of dilution which occurs. Thus, in charac-
terizing water flow, it  is important to consider the detail of the
flow pattern as well as the bulk  flow. Turbulent diffusion and
participate settling cause the transfer of material across flow  lines.
  For lakes and oceans, the ultimate distribution is governed
primarily by vertical transport. The  currents leading to this trans-
port depend on several factors relating to stratification, primarily
the temperature profile as a function  of the time of year. Other
phenomena which affect the distribution of containments in stag-
nant bodies of water are wind driven  currents, waves, turbulent
and eddy-thermal diffusion and particulate settling. The increase
in the solubility of gases with increasing pressure affects diffusion
and can lead to the trapping of gases in the deepest part of stag-
nant lakes.
  The transformation of a chemical species into one or more  other
species is also important in determining environmental fate. A key
transformation process  for organic pollutants is aqueous photolysis,
the alteration of a chemical species due to the absorption of  light.
Both direct photolysis, in which the pollutant itself absorbs the
solar radiation and sensitized photolysis,  in which the energy  is
transferred from another chemical  species, may occur.

Transport and Transformation in Soil
  The soil and rock material near the earth's surface contains
interconnected pore spaces that allow circulation of air and  other
gases and the transport  of either water and associated contaminants
or of liquids which are immiscible with water. Both liquids and
vapors can diffuse through the soil. This transport is characterized
by the time-dependent diffusion equation. Soils vary greatly in the
total pore space available, depending on composition and  com-
paction. Pores are classified as macropores, which allow the ready
movement of gases and percolating water or micropores, which
greatly impede gas movement and restrict water movement to slow
capillary action.  The  effective  diffusivity is depended on the
molecular diffusion coefficient, the soil porosity and the tortuosity,
a parameter which accounts for the ratio of the average path length
between two subsurface points to the straight line distance between
the points.
  When a liquid is in contact with the soil, a number of phenomena
occur. The liquid coats the pore walls  and particle junction  sites.
Evaporation of the liquid occurs at  interstitial soil surfaces, with
the vapor diffusing upward through the pores toward the soil-air
interface. Vertical liquid movement takes place through capillary
action,  which becomes the dominant  mechanism in wet, coarse-
textured material. For liquids which  are immiscible with water, the
liquid density plays a critical role. Liquids which are denser than
water will penetrate and settle at the bottom of the water table,
while lighter liquids will float on top of it. Adsorption also  takes
place and may be a dominant phenomenon in a dry medium.
  Diffusivity is a strong function of moisture saturation of the soil.
For moisture contents exceeding 40%, advection takes on parti-
cular importance.  Water movement in the upper layers of the soil
is influenced by the temperature and temperature gradients (Sovet
effect), while heat flow is influenced  by the movement of soil  water
(Dufour effect).
  Chemical biodegradation in soil is a significant environmental
process in the breakdown of organic compounds. The rate of bio-
degradation is a strong function of the organic matter content of
the soil.

Transport in Air
  The transport of vapors and particulates in the air may be divided
into two layers,  near the surface and far from the surface. The
dominant phenomena are different in the two zones. Transforma-
tion also can occur in the atmosphere, but the rates are relatively
slow and will not be treated here.
  The transport ofa. vapor from a soil  or liquid source into the
atmosphere is governed by both the movement of the air mass and
by diffusion. The important factors governing the transport rate
are the vapor pressure at the interface, the windspeed, the source
area and the roughness of the  surface. The atmospheric stability
and the molecular diffusivity  have a smaller effect.
  The movement of particulates from a  solid surface  depends on
the same atmospheric parameters as above  (windspeed, surface
roughness and stability). In addition, the entrainment of parti-
culates depends on the size, density and adhesion of the particles.
  In the atmospheric region above the surface layer, the transport
of both vapor and particulate matter is governed by the movement
of the air mass and by turbulent diffusion. Both the vertical tem-
perature profile, which determines stability and the windspeed are
important. Precipitation will  scavenge  particulates  and soluble
vapors from the atmosphere and gravitational settling will act to
redeposit particles on the surface.

Transport Across the Soil-Water Interface
  In addition to the interface between groundwater and the soil
which contains it, there are two  other types of soil-water interfaces;
soil erosion from rain and the interface between surface water and
the underlying bed. Erosion depends on the precipitation rate, the
configuration of the lands and the condition of the surface. The
interaction of a body of surface water with its bed is complex, with
the dominant processes being sedimentation, suspension and diffu-
sion. The primary factors governing the first two processes are
particle size and  density, while  the rate of diffusion is determined
by the mass transfer coefficient and the diffusivity.

Transport Across the Water-Air Interface
  The balance of a contaminant across an interface between water
and air generally can be modeled using Henry's Law, which makes
the Henry's Law constant for a given contaminant the most im-
portant factor for modeling this interface.  Since Henry's Law cons-
tant is a function of temperature and of the concentrations of any
other materials in the water, these secondary factors must be con-
sidered. It also is  important to realize that  Henry's Law was derived
for a static equilibrium condition and that the diffusion of the con-
taminant into the water and air phases can be the dominant factors
in determining the rate of transfer of the contaminant across the
interface.

Transport Across the Soil-Air Interface
  Communication of vapors between soil and air is really the com-
munication of the free air with the air in  the soil pores. The com-
position will be the same within the pores immediately below the
ground surface as it is immediately above the ground surface. The
rate at which a vapor from subsurface contamination is released
to the air thus depends on the rate at which  it evaporates below
the surface, the rate at which it diffuses to the surface and the rate
at which it is entrained into the atmosphere from the  surface, all
of which are related by the interface concentrations.

Transport Across the Water-Biota and
Soil-Biota Interfaces
  The details of  the transport of chemical species across the inter-
faces between soil or water and biological organisms are not com-
pletely understood. At present, probably the best way to model
these phenomena is to look at the equilibrium ratio between the
concentrations in a given species and in the surrounding environ-
ment. These ratios are known as partition coefficients. While it
                                                                                                        MULTI-MEDIA    477

-------
                            Table 1
   Importance of Phenomena in Various Exposure Assessment Stages
                       Table 2
Importance of Factors In Various Exposure Assessment Stages
Transport and Transformation
Phenomenon
Dissolution
Evaporation
Suspension
Sedimentation
Adsorption
Precipitation
Soil erosion
Henry's Law
Phase Equilibrium
Flow
Diffusion
Gravity
Stratification
Wind driven currents
Waves
Percolation
Capillary action
Advection
Sovet effect
Dufour effect
Liquid coating
Photolysis
Biodegradation
Water
0
U
0
i
1
2
2
(J
0
i
i
i
'I
:
t
u
0
i
0
u
0
^
i

Soil
2
U
U
0
2
2
2
0
0
i
2
2
2
0
U
i
2
i
i
2
2
U
2

A1r
0
0
0
0
1
2
0
0
0
2
2
2
2
0
0
0
0
2
0
0
U
1
0
Phenomenon
Ulssolutlon
Evaporation
Suspension
SedlnenUtlon
Adsorption
Precipitation
Soil erosion
Henry's _a.
Phase CqulHorljB
Flo.
Diffusion
Gravity
Stratification
Hind driven current
waves
Percolation
Capillary action
Ad>ect1on
Sovet effect
Dufour effect
Liquid coating
Photolysis
•lode gradation
So11-ซater
2
U
2
2
ฃ
2
t
0
i
t
2
2
t
0
0
i
2
(1
u
u
2
0
0
Soil -Air
t
2
2
U
2
2
U
2
2
'I
2
2

u
u
0
u
1
u
u
u
u
0
•ater-tlr
2
2
0
U
2
2
0
2
t
2
2
2
2
2
2
0
U
1
0
0
0
D
0
Soil-Biota
U
0
u
0
u
o
0
0
2
1
U
0
u
u
0
D
0
0
U
0
U
0
0
Kater-Blota
0
0
0
0
0
0
0
0
2
0
0
0
0
0
0
u
D
0
0
0
0
0
0
   2 • very I^Mrtant to BOdel
   I • Moderately laportant to Bode)
   0 • Unlaporunt to •ooel
is not feasible to experimentally determine partition coefficients
for all combinations of chemicals and biological species, it gener-
ally has been found that the octanol-water partition coefficient is
a good predictor of absorption in animal fat. For this reason, the
octanol-water partition coefficient has been widely used to esti-
mate other partition coefficients.

Determination of Human Exposure Levels
  It often is important to determine the level and type of human
exposure. There are three possible pathways: inhalation, ingestion
and dermal.  For inhalation, the level of exposure is determined
by the air  concentration, the breathing rate  and  the exposure
duration. For ingestion, the level of exposure is determined by the
concentrations in water and food the consumption rates for con-
taminated water and food and the exposure duration. For dermal
contact, the pertinent parameters are area covered by the material,
the concentration and the duration of the exposure.

CONCLUSION
  While the field of exposure assessment is complex, it possesses
a definable structure. This  general structure provides a basis for
Factor
Cnealcal Hum
Composition
Formula
Structural Formula
Holtcular might
Iflltlll Conctntratlon
Solubility
Density
Vapor Pressure
Henry 't law Constant
Utftusl.lv
until on
Particle Sllf
taount Released
Release Rate
Release location
Reltast Area
Reltale Temperature
Aaoltnt Temperature
Ttmpereturt Prom*
do. iutt
flow Pattern
fnotolysls Rate
tlodegradatlon Rate
Soil Porosity
Tortuosity
Nollture Saturation
Surface Roughness
Topography
•Inotpeed
AtJNltpnerlc jtis'lH/
PrtclplUtlon Rate
Breathing Rate
Consumption Rate
Contract Area
tiposure Duration
Cneracttrimlon
Source
1
I
i
2
t
2
/
2
I
I
I
I
1
0
0
0
u
a
u
u
a
0
u
u
0
u
0
0
0
u
0
0
0
0
u
u
Release
0
0
u
0
0
0
u
0
0
0
0
0
0
i
i
I
z
I
U
0
u
0
u
a
0
u
u
u
u
u
0
u
u
0
u
0
Transport
Biter
0
2
0
0
1
1
2
2
2
0
I
0
2
2
2
2
U
0
2
2
2
2
2
2
0
U
0
2
2
1
0
2
Q
U
0
U
0
and Transformation
Soil
0
2
0
0
1
1
2
2
2
0
2
u
2
2
2
2
0
0
2
2
2
2
0
0
2
2
2
2
2
0
0
2
A
0
0
0
0
Air
0
2
0
0
2
2
1
1
2
0
2
0
2
2
2
2
0
0
2
2
2
2
2
2
0
0
0
]
2
2
2
2
A
0
0
0
0
Factor
Cnralctl Name
Composition
Formula
Structural Formula
Molecular Weight
Initial Concentration
Solubility
Density
Vapor Pressure
Henry 'ซ Lax Cons tint
Deffuslvltj
Adneslon
Particle Mie
Amount Released
Release Kite
Release Location
Kelease Area
Melease Temperature
Ambient Temperature
Temperature Profile
Flew Kate
Flow Pattern
Photolysis Rate
bModtgradatton Kate
Soil Porosity
Tortuosity
Moisture Saturation
Surface Roughness
Topography
Wlndspeed
Atmospheric Stability
Precipitation Rate
Partition Coefficient
Breathing Rate
Consumption Rate
Contract Area
Exposure Duration
Exposure Pathway
Innalatlon Injestlon
0
0
0
0
0
0
u
0
u
0
0
0
0
0
u
u
u
0
u
0
u
0
0
0
0
u
0
0
u
u
u
u
0
z
0
u
t'
0
0
0
0
0
0
0
0
0
0
u
0
u
0
0
0
0
0
0
0
u
0
0
u
0
0
0
0
0
u
0
0
0
0
2
0
2
Dermal
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
u
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
2
478     MULTI-MEDIA

-------
Factor
Chemical Name
Composition
Formula
Structural Formula
Molecular Height
Initial Concentration
Solubility
Density
Vapor Pressure
Henry's Law Constant
Diffusely
Adhesion
Particle Size
taount Released
Release Rate
Release Location
Release Area
Release Temperature
Ambient Temperature
Temperature Profile
Flow Rate
Flow Pattern
Photolysis Rate
Bfodegraoatlon Rate
Soil Porosity
Tortuosity
Moisture Saturation
Surface Roughness
Topography
wlndspeed
Atmospheric Stability
Precipitation Rate
Partition Coefficient
oreatnlng Rate
Consumption Rate
Contact Area
Exposure Duration

Soil -Water
U
2
0
0
1
2
i
2
U
0
2
2
2
2
2
2
2
1
2
1
2
2
U
U
•i
'i
2
]
I
U
0
2
0
U
U
0
0

So1l-
Q
2
o
o
2
2
2
2
'i
2
2
2
2
2
2
2
2
2
Z
1
2
2
0
U
2
2
2
2
2
2
1
2
0
U
0
U
0
Transport
Air Water-
0
2
Q
0
1
2
2
2
2
2
1
0
U
2
2
2
2
2
2
1
2
2
0
0
0
U
0
0
0
2
1
'I
0
0
0
0
0
Across Interface
Air Soil -Biota
o
2
Q
o
1
o
2
U
0
0
0
0
0
0
a
Q
0
0
1
0
2
2
0
0
0
0
0
0
0
0
0
1
2
Q
0
0
0

Water-Biota
Q
2
o
o
1
o
2
o
0
0
0
0
0
0
0
0
0
0
1
0
1
1
0
0
0
0
0
0
0
0
0
0
2
0
0
0
0
Key:

2 • acre Man a factor of ten effect
1 ซ less tnan a factor of ten effect
0 ซ no effect
the organization of the phenomena and factors involved. The
representation of the relationship of various phenomena and factors
to the analysis of transport and transformation in various media
and across different interfaces in matrix form makes it possible
to determine in  the early stages of an exposure assessment what
data are necessary and  what phenomena need to be considered.

ACKNOWLEDGEMENT
  The authors gratefully acknowledge the financial and profes-
sional support of the Exposure Assessment Task Group of the
Chemical Manufacturers Association.

REFERENCES
1.  Thibodeaux, L.J., Chemodynamics, Environmental Movement of Chem-
   icals in Air, Water and Soil, John Wiley & Sons, Inc., New York, NY,
   1979.
2.  Neely, W.B. and Blau, G.E. Environmental Exposure from Chemicals,
   Vol. I, CRC Press, Inc., Boca Raton, FL,  1985.
3.  Neely, W.B. and Blau, G.E. Environmental Exposure from Chemicals,
   Vol. II, CRC Press, Inc., Boca Raton, FL, 1985.
4.  Rosenblatt, U.K., Dacre, and J.C. Cogley, D.R. "An Environmental
   Fate Model Leading to Preliminary Pollutant Limit Values for Human
   Health Effects." in Environmental Risk Analysis for Chemicals, Chapter
   15, pp.  474-505, R.A.  Conway, Ed., Van Nostrand Reinhold Com-
   pany, New York, NY,  1982.
5.  Gilford, J.H., "Environmental Effects Assessment of New Chemicals
   under the Toxic Substances Control Act," AICE, 1985 Summer Meeting,
   Paper 37B Seattle, WA, Aug. 25, 1985.
                                                                                                        MULTI-MEDIA    479

-------
                       The  Use of  Stabilized  Aqueous Foams  to
                                   Suppress  Hazardous  Vapors —
                    A  Study of Factors  Influencing  Performance

                                                   Roger R. Aim
                                                Chris  P. Hanauska
                                                 Kathleen A. Olson
                                              Myron T.  Pike,  Ph.D.
                                                  CH3M  Company
                                      Industrial  Chemical Products Division
                                                St. Paul, Minnesota
ABSTRACT
  During remediation of sites containing VOCs chemically con-
taminated soil and solid wastes, a convenient vapor suppressive
cover is needed to maintain the level of volatile organic compounds
(VOCs) within acceptable limits. To answer this need,  3M has
developed a stabilized aqueous foam system which is easily applied
as a fluid foam using standard foam equipment but within minutes
turns into a tough, elastomeric, non-draining foam.
  The effectiveness of stabilized foam as a vapor suppressing
medium is influenced by foam variables such as formulation, foam
depth, expansion ratio and age, as well as the nature of the particu-
lar hazard. This paper presents results of laboratory and field tests
which have been conducted  to investigate the influence of these
factors. The paper also includes detailed information about the
equipment and test procedures utilized in the laboratory for foam
preparation and evaluation.  Based on these test results and some
theoretical modeling, an attempt has been made to draw conclu-
sions concerning the primary means of VOC molecular transport
through the foam layer.

INTRODUCTION
  Remediation of hazardous waste sites has begun in earnest due
to the billions of dollars authorized by SARA. Most of the attention
has focused on the detoxification and/or approved site storage of
hazardous waste to prevent groundwater pollution problems. In
excavating and transporting this waste, another pollution problem
must be addressed—that of air pollution, as it may affect both site
workers and residents of adjacent communities. Existing methods
used to maintain air quality during remediation, such as coverage
with dirt, kiln dust, tarps or polymer films, can mitigate vapors
to varying degrees; however, they are all labor intensive and often
add to the burden of material to be ultimately removed or treated.
Additionally, they cannot be applied immediately while  waste is
being disturbed.
  Stabilized aqueous foam systems have been developed by 3M
which can be conveniently applied using commercially available
air-injecting or air-aspirating foam equipment. Proprietary foaming
and gelling agents are premixed, educted or injected into a water
stream producing a very fluid foam which changes within 1-4 min
to a tough, elastomeric, non-draining foam. Over several days, this
stabilized aqueous foam will dehydrate and collapse to an elasto-
meric membrane which retains excellent consolidation  and vapor
suppression properties.
  Laboratory, field and environmental test results using these
foams have been described in previous papers1'2. This paper will
attempt to better quantify foam and hazard variables affecting
vapor suppression performance.

EXPERIMENTAL
Laboratory Standard Test Procedure
  Figure  1 shows the laboratory test apparatus used by 3M to
evaluate vapor suppression performance of stabilized foams; this
system is patterned after the flux chamber designed and employed
by Radian Corporation to monitor emissions at hazardous waste
sites3  In a typical experiment, 5270 g of clean 100 mesh sand are
placed in  the bottom portion of the 5 I, 25 cm diameter sampling
chamber along with 920 ml of test organic liquid, which is enough
liquid to totally wet but not submerge the sand. The top portion
of the chamber is then mated to the bottom through a ground glass
joint around the circumference. The stirrer is then started (approxi-
mately 2 rev/sec) and a nitrogen sweep of 3.6 1/min is allowed to
flow through the chamber, picking up \ olatile organic compounds
(VOCs) emitted from the solvent-saturated sand. This sweep is then
split, with the minor (approximately 3^o) portion directed through
the gas sample loop of a Hewlett-Packard 5-90 gas chromatograph
(GC) equipped with a  flame ionization detector (FID) and a 3393A
computing integrator; the rest of the gas flow is directed out of
the system through the vents. Samples (0.25 ml) from this sample
loop are analyzed automatically at  regular intervals (from every
5 min to every hour depending on the rate of change in emissions
produced and GC analysis time). Analysis of each sample appears
as a series of peaks on the chromatogram produced by the integra-
tor. Each peak represents a different chemical component of the
VOC and exhibits a unique retention time on the GC column. The
area of each peak, which is proportional to the quantity of material
present in the 0.25 ml sample, is read directly from the chromato-
gram as area counts. Several samples are taken to establish a pre-
foam  emission rate.
  After completion of baseline measurements, the tubing leading
from  the  flux chamber to the GC gas sample loop is swept with
nitrogen.  The upper portion of the chamber is removed, wiped with
tissue and swept with nitrogen. During these procedures, the lower
portion of the glass vessel is covered with a glass plate to prevent
drying from the surface of the sand/solvent slurry. Sampling is
continued during the nitrogen sweep to determine when the line
is completely clear. Foam is prepared by one of the following two
methods,  depending upon the desired expansion ratio (foam volume
to liquid  volume).
  For low expansion ratio foams (4:1), the foam solution (300 ml)
is prepared by quickly mixing 18.0 ml of FX-9161 stabilizer with
18.0 ml FX-9162 foamer premixed in 264 ml tap water. This solu-
tion is immediately transferred to a large (approximately 41) com-
480    MULTI-MEDIA

-------
mercial Waring blender and foamed for 15 sec at low speed. Then,
as quickly as possible, the foam is poured onto the VOC saturated
sand, yielding 0.62 g/cm2 application weight. Within 1 to 2 min
the fluid foam turns into a tough, elastomeric, non'draining sta-
bilized foam.
  For higher expansion ratio foams (18:1), 150 ml of the same test
solution is foamed using a Hobart N-50 commercial food mixer
using a wire whisk beater and mixing for 15 sec at high (#3) speed.
The  foam is quickly transferred to the chamber, yielding 0.31
g/cm2 application weight. The chamber is again closed and peri-
odic emission readings are taken as before using the same 3.61/min
nitrogen sweep. With foam in place, there is headspace of 1.5 to
1.8 1 in the sealed vessel, resulting in a residence time of approxi-
mately 0.5 min for the nitrogen sweep flow.
  Experiments with foam cover were  run for at least 24  hr and,
in one case,  up to 12 days. After completing these measurements
with foam cover, the chamber was opened, the foam cover was
removed and the chamber was reassembled to make a second VOC
baseline reading. This was done to determine whether the VOC-
saturated sand had lost its vapor generating capabilities over the
extended duration of the tests; if the baseline reading had become
significantly lower than the originally measured values, the data
points gathered at extended times should be considered with this
depletion of source material in mind. All tests were run at ambient
laboratory temperature, which was 22 ฑ  1 ฐC.

Data Analysis
   The number of area counts recorded from the FID is  propor-
tional to the number of molecules of VOC coming through the
GC column. The GC can be calibrated by using standards of known
concentration. This calibration is used to establish the concentra-
tion, in atmospheres of partial pressure, of the VOC in the nitro-
gen sweep flow from the chamber. This concentration can be used
to calculate the  mass flux from the surface of the foam using  a
mass balance around  the chamber headspace (Figure 1):
                                                                 T    = Ambient temperature;
                                                                                                      K
             J =  (p,*V*MW) / (A*R*T)

J    =  Mass flux;    —|—
                     cm2 sec
Pi   =  Partial pressure of VOC;    atm

V   =  Volumetric nitrogen flow rate;     ^~

MW =  Molecular weight of VOC;    —iy-
                                    mole
A   =  Surface area of foam;    cm2
                                                        (1)
                             MOTOR AND
                             STRIflRER
\' ' ' ' S SB / I
\ HAZARDOUS MATERIAL 2/ //
Y / / / / 7/1

QA9 CHROMATOORAPH
WITH FLAME IONIZATION
DETECTOR
                           Figure 1
          Flow Scheme for Vapor Suppression Apparatus
                                                                 The flux, J, of VOCs through the foam can be modeled at steady
                                                               state using a form of Pick's 1st Law of Diffusion4:
                                                                                    T _ (P2 ~ Pi)
                                                                                          (h/P)
                                                       (2)
                                                                 P  = Effective permeability of the foam;

                                                                 h  = Depth of foam layer;    cm
                                                                                                                g
                                            cm  sec atm
                                                                 p2 = Partial pressure of the VOC at lower foam surface;  atm

                                                                 p, = Partial pressure of the VOC at upper foam surface;  atm

                                                                 In this equation,  (h/P) is the  resistance and  (p2-p,) is the
                                                               driving force for the diffusion flux J. A number of things can be
                                                               inferred from this equation:

                                                               • Increasing the depth, h, of the  foam layer would increase the
                                                                 resistance to diffusion

                                                               • Reducing the  partial pressure of the VOC below the foam, p2,
                                                                 would lower the driving force

                                                               • Decreasing the permeability,  P, of the foam will increase the
                                                                 resistance to diffusion and thus reduce the flux; the permea-
                                                                 bility of the  foam may change with:  foamer  concentration,
                                                                 stabilizer concentration, expansion ratio and the properties of
                                                                 the VOC,  such  as solubility in  the foam

                                                                 The foam's effectiveness often is presented in terms of the per-
                                                               cent suppression—a comparison of the flux rate with foam to the
                                                               baseline flux rate. It is calculated  from the formula:
                                                               T,      o       •      A    flux rate with foam cover \ „ , Mm
                                                               Percent Suppression =(1	-.	-.	=	:	 |x 100%
                                                                                                                         (3)
                              baseline flux rate


LABORATORY RESULTS
Effect of Volatile Organic Compound Tested
  In Figure 2, the results of three experiments are combined to
show the effectiveness of 0.62 g/cm2 of 4:1 expanded  foam at
suppressing vapor emissions of sand saturated respectively with
cyclohexane, acetone and methyl ethyl ketone (MEK).

                     Volatile Organic Compounds In Sand
45.0 •

40.O •

35.0 •
30.0 •

25.0 •
20.0 -

15.0 -

10.0 -
5.0 -
0.0 -

g
o


*









D Acetone
o Cyclohexane

>+ Methyl Ethyl Ketone

Initial Values,
Pre— Foom %
/
'
Final Values,
s^ Foom Removed
D ^ \^^
o ^^-^



                                                                                                   30

                                                                                                Time (hours)
                                                                                         Figure 2
                                                                              Stabilized Foam Vapor Suppression
        = Ideal gas constant;
                             cm3 atm
                             moleฐK
  Using the standard test procedure described in the Experimen-
tal section, nearly total (>99.9%) vapor suppression was observed
with cyclohexane for the first 2 days. Even after 12 days, when
                                                                                                      MULTI-MEDIA    481

-------
the foam had dehydrated to form a thin, tough, elastomeric mem-
brane, vapor suppression remained over 97%. After removing this
membrane along with 0.6-0.9 cm of consolidated sand which ad-
hered to it, emission tests were run again. Cyclohexane vapors were
emitted at close to the original baseline level, indicating that the
potential for emissions had not diminished appreciably.
  Vapor suppression results were significant, though less spec-
tacular, when the ketones were evaluated using the standard test
procedure.  Over  the first 24  hr, compared to initial  baseline
emissions, percent suppression for both MEK and acetone remained
around 85%. However, due  to gradual  evaporation, the ketones
showed depletion from the uppermost layers of sand. Measure-
ments made after foam removal indicated a 50% reduction of MEK
emissions and a 60% reduction of acetone emissions, compared
to the pre-foam emission rates. The loss of emission potential con-
tributes to the leveling and decreasing of the emission rates in the
later parts of the experiments.
   Solubility of the VOC in  water is an important parameter in
predicting vapor suppression effectiveness of stabilized foams, as
the hydrophilic polymer phase of the foam provides the greatest
resistance to diffusion. With  both ketones, flux increased steadily
over the first few hr as the  concentration of VOC in the foam
increased.

Effect of Foamer and Stabilizer Concentration
   FX-9162 foamer and FX-9161 foam  stabilizer were evaluated
at higher and lower concentrations than their usual 6% to deter-
mine effects on vapor suppression. Using the standard test pro-
cedure, results of these variations for cyclohexane are shown in
Figure 3. In this figure and in subsequent  figures, the graph was
plotted using foamed values only—the pre-foam values being off
scale. As shown in Figure 3, varying stabilizer concentration from
6% to 12% had little effect  on emissions. Varying foamer con-
centration from 3% to 6% to 12% slightly increased the emissions
with each experiment. However, suppression was greater than 99%
for all experiments.
                           CycloheKOne In Sond
                                                                                                Liquid Acelone




1
\
u
CT

O*
9

0

1




1S.O -
15.0 -
14. 0 •
13.0 •
12.0 •
1 1.0 •
10.0 •

9.O •
8.0 -
7 O •

8.0 •
5.0 -
4.0
3.0 -
2.0 -
l.O -
a ft -.

Pre-Foom Fluซ. -10.OOO (g/eq cm/min): *5

o00oooaฐฐ .ปซ.****
oฐฐ0ฐฐOODฐ 	 ******
aDoฐฐฐ 	 ..'*•*


rf*
jfT ....••••*
ff* •••
JC ••••*
f .••0<"
ฃ ^ 	 *
n n~*
ฃ
T ซซ a *% S(obii.ier
•" .'
tt • * 6X Stobiii*ซr
* oo
b^ฐ ซ 12* Stooi it*
tr
                                                                                                 '2      18      2O     2ซ     28

                                                                                                T]~ซ (houre)
                                                                                            Figure 4
                                                                                   Varied Stabilizer Concentration
                                                                  Effect of Application Weight
                                                                    The effect of difference in foam application weight on vapor
                                                                  suppression is shown in Figures 5 and 6. In each experiment, the
                                                                  foam expansion  ratio was 4:1.

                                                                                             Cydoheซa~ซ In Send
                                                                  5
                                                                  >
0.015 •
O.O14
0.013
0.012 -
0.011 -
0 01O -
0009
0.008
0.007 -
O.O06 -
0.005 -
0.004
O.O03 -
0.002 -
0.001 -
9 Pre-Foom Fluซ. '10.0OO (g/ซq cm/mln): 31
o + 3% Foomer. 6% Stabilizer
0 a 8X Foamer. 6X Stabilizer
e 12% Foamer. 6% 5tobzer
* D 6X Foomer. 12% Stab^zer
"•....-......•••'
O
0 DD0ฐ0ฐ0
***•*• ฐ. '•••..* ....•• .

O.05
O.OO -
C
-, ซ.o -
c
ฃ
E'
0 4 8 12 18 20 24 <ฃ 3.0 •
Time (houri) JJ
Figure 3 I 2 ฐ
Varied Foamer, Stabilized Cone. "
3
..0 -
n evaluating foam composition effects on acetone, the test
cedure used was slightly different, employing approximately 0 0 _,
0ฐ
a
0 	 J
ri__ 	 _„_ a --niimfl ccooocooo^^
20 *O 6O BO 1OO 120 140 1 60 ' 8C 2O
rmm (hour*)
Figure 5
Varied Foam Application Weight
A.cซtonซ In Son<]
Pre-Foom Flu*. • 1 O.OOO (fl/tq cm/rr>in): 33
* 0.31 Q/ซq cm
4>>****^^ o 0.62 q, BQ cm
* * o ^ .2* g.'tq cn>
ฐ ฐ ฐ A ซ *
•••-•ป
* ' *ฐ'^.
/ ••'
+ o
4-
* o o a o o ฐ o a o o o o a o o a a o o o o a o oa o o
^0ฐฐฐฐฐฐฐฐฐ
2.5 1 of acetone in the test chamber without any sand. Figure 4
shows that varying the stabilizer concentrations from 4% to 6%
to 12%  reduces the acetone emissions. No foamer concentration
variations were evaluated with either acetone or MEK.
                                                                                        1O           20
                                                                                                Time (hours)
                                                                                             Figure 6
                                                                                   Varied Foam Application Weight
482    MULTI-MEDIA

-------
  Figure 5 presents data for sand wet with cyclohexane. In the first
24 hr, the  0.62  g/cm2 application weight allowed 60%  of the
emissions of the 0.31 g/cm2 application weight. At 24 hr, the rate
of emissions for the lower application weight experiment increased
dramatically; the same effect was noted with the higher applica-
tion weight experiment at 170 hr  (7 days). We believe this abrupt
change in flux rate occurs when the foam becomes completely
dehydrated and thus increases the permeability of the foam.
  After the rapid change in  emission rate was observed,  the
experiments were disconnected from the GC, but the sweep flow
was continued. After about 290 hr (12 days) from the initial foam
application, the experiments were reconnected to the GC. The emis-
sion rate for the higher foam application weight was 75% that for
the lower application weight.
  Despite the variations  in flux between these foam application
weights and the differences  over time, the percent suppression
remained greater than 97% compared to the prefoamed and foam-
removed flux  rates.
  Application weight  also  proved  to be an  important  factor
affecting emissions from acetone in sand, as shown in Figure 6.
Varying the application weight from 0.31 to 0.62 to 1.24 g/cm2
increased the percent suppression at 10 hr from 78% to 81% to
93%  respectively.

Effect of Foam Expansion Ratio
  The effect of differences in foam expansion ratio on vapor sup-
pression is shown in Figures 7 and  8. The expansion ratio was varied
from 4:1 to 18:1, keeping the application weight at 0.31 g/cm2.
Figure 7, presenting flux data for sand wet with cyclohexane, shows
the emission rate using 18:1  expanded foam was 50 times that of
the 4:1 expanded foam. However, the percent suppression for both
18:1 and 4:1 expansion was  greater than 99%. Figure 8 presents
flux data for sand saturated with acetone. Early in the experiments,
the 18:1 expanded foam  suppressed emissions better; after 7 hr,
the 4:1 expanded foam performed better. However, the difference
in emission rates was never  greater than 15%.

                            Cyclohexane  tn Sand

0.45 -
0.40 -
0.35 -
0.30 -
0.25 -
0.20 -
0.15 -
0.10 -
0.05 •

o.oo ^
Pre— Foam Flux, *1 0,000 (g/sq cm/min): 31
+
•f

/
+ D Expansion Ratio of 4: 1
+ -f Expansion Ratio of 18:1
t-
h
a
*• D
- D DDDD na n nann DO na D noo a ooa a
                                   20

                               Time (hours)
                             Figure 7
                   Varied Foam Expansion Ratio
 FIELD TESTS MEASURING VAPOR
 SUPPRESSION ABILITY OF STABILIZED FOAM

   In March 1987, Radian Corporation evaluated the performance
 of stabilized foam for suppression of vapors from soil contami-
 nated with JP-5 fuel at the Marine Corps Air Ground Command
 Center in 29 Palms, California5. This location,  in the southern
 Mojave desert, had severe weather conditions (hot, dry, windy and
 sunny) during testing. Initial, uncontrolled hydrocarbon emissions
 measured from this soil using the Radian flux chamber indicated



c
E
X
u
?
5


o
•
X
3
C






1 1.0 -
10.0 -
9.0 -
8.0 -

7.0
6.0 -

5.0 -

4.0
3.0
2.0
1.0 -
00
c

Acetone In Sand
Pre-Foam F\ux, *1 0,000 (g/sq cm/min): 35
+ Foam Expansion Ratio of 4:1
u Foam Expansion Ratio of 1 8:1 0
D
/Y
\ +
\
Fnal Values. \ 	 > +
Foam Removed +
n D u +
D D ฐ D D
+ nฐฐ +++++*+ D0D
a + + ^
^ฐu ++++++++++*
j?
3 10 20 30 4
Time (hours)
                           Figure 8
                 Varied Foam Expansion Ratio

they were C7 thru C10 hydrocarbon species (28% paraffins, 56%
olefins, 5.3% aromatics and >0.5% each of halogenated and oxy-
genated compounds).
  The following procedure was used to run a typical vapor sup-
pression test. A circular test pad roughly 25 cm thick and 1 m in
diameter was formed from freshly excavated contaminated soil.
The flux chamber was then placed on the soil to obtain a baseline
flux reading. The flux chamber was removed and stabilized foam,
prepared by educting 6% each of FX-9162 foamer and  FX-9161
foam stabilizer  into a  pressurized water  stream and  foaming
through an air-aspirating nozzle, was applied at 2 to 5 cm of depth
on the test pad. After the foam had gelled, the flux chamber was
placed back  on the pad and flux readings were taken periodically
for the duration of the test. At the end of each experiment,  the
foam (with some consolidated sand) was carefully removed from
the soil and uncontrolled emissions were again measured to deter-
mine whether the test pad retained its emissions potential (less than
70% of original baseline flux meant that readings with foam were
considered invalid).  Percent VOC control or percent vapor sup-
pression, was then calculated using the original baseline reading
as a standard.
  The test results, presented in Table 1, show the performance of
stabilized foam applied at: (1) a low expansion ratio (5:1) on a
horizontal test pad, (2) a medium expansion ratio (18:1) on another
horizontal test pad and (3) a low expansion ratio (5:1) on a non-
horizontal surface (a 2:1 slope). Better results were achieved using

                           Table 1
                 Summary of Field Test Results
                                                                    Estimated Foam
                                                                    Expansion Ratio
                                                                       7:1
               Test Pad
              Orientation
                                                                                    Flat
 Test
Duration
                                                                                             7 days
                                                                                             6 days
 Post-Foam
Emissions Check
                                                                                                        Acceptable
                                                                                                      Not Acceptable*
    7:1
               2:1 slope
                          25 hours
                                     Acceptable
Summary of
 Results
                                                                                                                    99% VOC Control
                                                                                                                      for J days
                        lOOt VOC Control
                           for 1 day

                        901 VOC Control
                           for 3 days

                        lOOt VOC Control
                           for 6 days


                        90t VOC Control
                          for 25 hours
* Post-foam Missions potential ซas <70t of Initial potential.


                                     MULTI-MEDIA     483

-------
5:1 vs. 18:1 expanded foam as would be predicted from the labora-
tory results presented earlier. The low expansion foam continued
to show  <99% suppression after 1  week, while  the  medium
expansion foam showed reduced vapor suppression after one day.
Nevertheless, the lofty character of the medium expansion foam
would make it useful for short-term applications involving cover-
age of highly contoured surfaces.
  Good results were achieved with the low expansion foam on the
2:1  sloped surface. Though the bulk of the foam layer could not
be maintained due to runoff prior to gelation, a thin soil/gel layer
of 0.3-0.6 cm thick remained. In all cases, the surface of the foam
gradually dried over several days to a tough leather-like, elastic
membrane covering a still moist foam which continued to exhibit
good  vapor suppression  characteristics for the  low expansion
foams. This foam/membrane hybrid also showed excellent soil con-
solidation performance, keeping the easily wind-blown desert sand
in place.
CONCLUSIONS
  Laboratory and Held tests were conducted with aqueous stabi-
lized foam to investigate the effects of foam variables and the
nature of the hazard on vapor suppression performance. The fol-
lowing trends were noted:

• For a period of days, the percent suppression of hydrocarbons
  did not change significantly. In a 12-day laboratory experiment
  with cyclohexane and a 7-day field trial with JP-5 fuel, the per-
  cent  suppression was greater than 97%, even after the foam had
  dehydrated to form a membrane.
• With high polarity VOCs, such as acetone and MEK, percent
  suppression, though in the 90-100% range for the first several
  hr, decreased to the 80-90% range after 10  hr for foam appli-
  cation weights of at least 0.62 g/cm2. The higher polarity allows
  these VOCs to diffuse faster than hydrocarbons through the
  aqueous matrix of the foam. (These percent suppression calcu-
  lations are based on initial, pre-foam emission values and do
  not account for depletion of the  VOC from the soil).
• In general, vapor suppression properties of stabilized foams were
  not greatly affected by variation in concentration of the FX-9162
  foamer  and  FX-9161  foam stabilizer  components.  Some
  improvement in suppressing  acetone vapors was noted when
  FX-9161 stabilizer concentration was doubled from 6% to 12%,
  while a slight decrease in suppression of cyclohexane vapor was
  noted when FX-9162 foamer concentration was increased.
• The application weight of stabilized foam used should be deter-
  mined by the nature of the hazard. Lowering the application
  weight of 4:1  expanded foam from 0.62 to 0.31 g/cm2 did not
  significantly  hurt performance on cyclohexane.  However,
  doubling the application weight of stabilized  foam from 0.62
  to 1.24 g/cm2 on acetone cut emissions by more than 50%.
• Vapor suppression performance was affected  by foam expan-
  sion  ratio,  particularly with nonpolar VOCs such as cyclo-
  hexane, based on both laboratory and field tests. Thus, increasing
  air content of foam  to improve coverage should be employed
  only after careful consideration.

REFERENCES
1.  Aim. R.R., Olson, K.A. and Peterson, R.C. "Using Foam to
   Maintain Air Quality During  Remediation of Hazardous Waste
   Sites," APCA Annual Meeting, June 1987.
2.  Aim, R.R., Olson, K.A. and Reiner,  E.A. "Stabilized Foam—A
   New Technology for Vapor Suppression of Hazardous Materi-
   als," Proc. International Congress on  Hazardous Materials
   Management, June 1987,  583-595.
3.  Schmidt, C.E. and Balfour, W.D. "Direct Gas Emission Meas-
   urement Techniques and the Utilization of Emissions Data from
   Hazardous Waste Sites," Proc. National Conference on Envi-
   ron. Eng. ASCE, 1983.
4.  Ostrogorsky, A.G. and Glicksman,  L.R. "Rapid, Steady State
   Measurement of the Effective Diffusion Coefficient of Gases
   In Closed-Cell Foams," Transactions of American Society of
   Mechanical Engineers, J.  Heal Transfer,  in  press.
5.  Radian Corporation, "3M  Temporary  Foam FX-9162  and
   Stabilized Foam FX-9161/FX-9162 Evaluation for Vapor Miti-
   gation at 29 Palms, CA" DCN 087-204-138-03, June 1987.
484    MULTI-MEDIA

-------
                        Multimedia Approach  to  Risk Assessment
             for Contaminated Sediments in  a  Marine  Environment

                                               Seong T.  Hwang,  Ph.D.
                                      U.S. Environmental Protection  Agency
                                            Exposure Assessment Group
                                                  Washington, D.C.
ABSTRACT
  Problems of contaminated sediments frequently occur when one
makes exposure evaluations of rivers, lakes, estuaries and harbors,
which represent a special type of Superfund sites. Contaminated
aquatic sediments can lead to significant risk to humans resulting
from exposure to contaminants through multimedia pathways such
as ingestion, dermal absorption and inhalation. This paper exa-
mines risk assessment approaches for these pathways under differ-
ent conditions and provides predictive models to estimate the
ambient air concentrations. Certain observed data are presented
along with predictions pertinent to the present no remedial alter-
native. Remedial alternatives are also examined. Combined risks
from the multimedia pathways can be obtained by considering the
activity patterns of the affected population once the media con-
centrations are known.

INTRODUCTION
  Contaminated waterways in a number of widespread locations
are drawing increased public attention. The public is concerned
about the health consequence of eating fish caught in harbors,
lakes, or estuaries with high levels of chemical contaminants in
the bottom sediments. The public is also concerned about exposure
from recreational activities  on  the waterways or the contiguous
land. Because of high concentrations of contaminants in fish In
some areas, waters are closed to fishing and signs can be found
in certain accessible places to warn the public about the potential
for health risks resulting from  water contact.
  These waterways with contaminated sediments are one type of
Superfund site, but they are unique in terms of the nature of con-
tamination and releases of the contaminants to which the public
can be exposed. Although their  numbers are limited compared to
the total number of Superfund sites, the potential for public health
risks from sediment contamination extends beyond the ingestion
of contaminated fish. Public health evaluations can be a compli-
cated process requiring multimedia monitoring data and assess-
ments of alternate disposal or treatment processes.
  In many of these sites, major contaminants of health conse-
quence found in sediments are PCBs which were mainly discharged
into the waterways by industrial manufacturing facilities handling
PCBs. PCBs constitute a mixture of biphenyl compounds with chlo-
rine substitutions in different positions on the ring structure. Other
chemical constituents also found in these sediments include metals,
dioxins and furans.
  The PCB concentrations  in sediments range from a few ppm
to many percent in some hot spots scattered over  the area. The
severity of contamination varies depending upon sample locations
in relation to the source of contaminant discharge and migration
behavior in the rivers, estuaries or lakes.
  To evaluate the potential of public health risks, responsible
parties  at  several  sites have been  continuously sampling  and
analyzing sediments, ambient air quality and aquatic and marine
species  as part of an effort to assess the public health risks asso-
ciated with existing contamination and remedial action alternatives.
It is also possible that terrestrial species are affected as a result
of a prey-predator relationship at different trophic levels or simply
by being exposed to the contaminated media.
  Despite an abundance of benthic microorganisms in the sedi-
ment layer, the contaminants of major concern such as PCBs are
highly resistant to biodegradation and will stay persistently in the
sediments.  Metals  will  not biodegrade.  These contaminants
routinely discharged from the manufacturing process or from spills,
would have settled slowly to the bottom to form sediment or be
incorporated into the sediments.
  The  flow of water in rivers and tidal fluctuations in oceans
influence the extent to which the sediments are migrated down-
stream  and the distribution of contaminants being carried by the
migrating sediments. Storm events have a major effect on the con-
taminant redistribution by resuspending the sediments. Hence, the
sediment and concentration distributions over the contaminated
areas could be a dynamic process which could change over time
under the influence of tidal flows, change in river flows, storm
events  and some man-created activities such  as passing boats.
  Although sediments will act as a reservoir and hold most of the
contaminants, the contaminants are slowly released to the water
column or to the air either through the water column or directly
from the exposed surface of the sediments when tides or water levels
are low. The slow release from one phase to another phase occurs
even though the contaminants are sparingly  soluble in water or
have extremely low vapor pressures.
  Hence, human exposures can come from sediment, surface water
and air. Although one major component of the health risk assess-
ments is the potential for bioaccumulation of the contaminants
in sediments and water by the aquatic food chain and subsequent
transfer to the human body through the intake of fish or shell-
fish, other routes  of potential exposures are  also important.
Exposures can also result from  recreational activities  such  as
swimming, fishing, sailing and picnicking. Swimming results in
dermal exposure upon contact with contaminated sediments and
water and also can result in ingestion; the contaminant released
into the air will also be inhaled.
  There are numerous pathways that need to be considered in
health risk assessment and in evaluating remedial actions. In dealing
                                                                                                  MULTI-MEDIA    485

-------
with cleanup alternatives, the long-term health consequence after
cleanup as well as the short-term effect of the elevated ambient
concentrations that can be created as a result of cleanup related
activities such as dredging, etc. should be included in the integrated
exposure assessment. The storage or disposal of dredged sediments
should also be a part of this assessment. This paper will address
some of these pathways and identify relevant factors influencing
the health risk assessment for the present no action situation or
when alternate remedial actions are considered.

EXAMPLES OF CONTAMINATED AREAS
  Several sites with a significant amount of PCS contamination
are publicized in the open literature  (2,9). A few are being evalu-
ated for the sediment reclamation project from the standpoint of
public health, environmental impact and feasibility.

New Bedford  Harbor
  Encompassing a geographic area which includes New Bedford,
Fair Haven, Dartmouth and Acushnet, the contaminated areas
extend  from the Acushnet  river throughout  the New Bedford
Harbor to the bay areas outside the hurricane barrier. Extensive
sampling efforts have been conducted by state and  Federal organi-
zations to determine the levels of contaminants in sediments, air
and eco-species. The highest levels of PCB  concentrations in
sediments are reported in the Acushnet Estuary area, and the sedi-
ment PCB concentrations range from 1 to almost 200,000 ppm.
In addition to PCBs. chromium, lead, dioxins and furans also are
found in the sediment samples. The ambient air concentrations in
the area range from  1  to  10 ng/m3 of PCBs for the  "back-
ground" area, to concentrations in excess of 100 ng/m3 in the
downwind ambient air. Former major users of PCBs include  two
electrical capacitors manufactures,  Aerovox Inc. and Cornell—
Dubillier Electronics Corp., who discharged PCBs to the estuary
and to the municipal wastewater treatment system from the 1930s
until 1977.
  The Acushnet Estuary and the New Bedford Harbor are closed
to all fishing. The bay areas south of  the hurricane barrier are open
only to certain species of fish. Bedsides the fish  ingestion path-
way, the number of populations affected by the contaminated sedi-
ments could be significant. The inhalation of volatilized or airborne
contaminants as well as dermal contact could be important path-
ways, because a significant number of industrial facilities  and the
general population are located along the coastlines of the estuary
and harbor. The contamination of the wastewater  treatment plant
located at the corner of Clark's Cove  that received wastewater from
the facilities manufacturing electrical capacitors has resulted  in a
widespread contamination of  sediments in the open bay area.

Hudson River
  The highest quantity of contaminant  sediments  is  located
immediately downstream of the former discharge points in the
upper Hudson River  above Troy, New York.  The PCB concen-
trations in the upper portion of the Hudson River are reported to
range from a few ppm to 1000 ppm with the low levels of PCB
concentrations in the center of the river and high  levels along the
depositional shore. As a result of sediment migration, PCB con-
tamination of sediments is found far downstream of the discharge
point. In the lower portion of the Hudson River, New York Har-
bor has the greatest mass of PCBs at an average PCB concentra-
tion of 3 ppm.4 It is estimated that about 75% of the PCBs in
New York Harbor originated from discharges  to the upper Hud-
son River.1
  In a law suit brought by  the New York State  Department of
Environmental Conservation (NYSDEC) in 1976, the General Elec-
tric  Company was identified  as the  primary source of PCB
discharge into the river. The PCB discharges continued for about
30 years from GE's two capacity manufacturing facilities.2
  Since 1976,  much of the Hudson  River is closed to fishing by
NYSDEC because the levels of PCBs in many fish exceeded the
tolerance level (Sppm) set by the U.S. Food and Drug Adminis-
tration. Milk from cows fed pasture forage grown in the Hudson
River region  was tested for the PCB content.4 The animal feed
contained PCBs in excess of the FDA limit of 0.2 ppm. but it is
reported that all eight samples of milk tested showed PCB levels
below  the FDA standard. NYDEC reported the results of air
monitoring around the Hudson River area. Ambient air PCB con-
centrations in 1977 are reported to range from 20 to as high as
several thousand ng/m3.

Waukegan Harbor
  The site is located near Waukegan Harbor about 37 miles north
of Chicago. Industrial facilities located around the contaminated
area include the Outboard Marine Corporation (OMQ, Larsen
Marine, a National Gypsum Plant, Falcon Marine and the Wauke-
gan Water Filtration plant. The U.S. EPA reported high levels of
PCB in water and sediments in the vicinity of OMC.7>> In some
areas, PCB concentrations as high as 25% in sediments have been
observed. Some hot spots exceeding 10,000 ppm (1 %) in PCB con-
centration are also identified.
  Various fish species inhabit the  Waukegan Harbor area. Fish
containing  PCB concentrations exceeding  100 ppm have been
caught within the harbor.' This level  considerably exceeds the
FDA tolerance limit of 2 ppm established in August of 1984. No
active  discharges  of  PCBs contaminants  into the harbor  are
reported at present.

Sheboygan Harbor, Wis.
  A Congressional bill recently has been introduced to mandate
the cleanup of the Sheboygan Harbor in Wisconsin in 3 years and
to provide a permanent disposal facility for the dredged sediments.
The bill also includes a 5-year study for a demonstration project.
focusing  on  the removal of "toxic sediments  from the lake
bottom."

Bloomington, Ind.
  The discharge from the city's wastewater treatment plant which
received PCB-contaminated wastewater resulted in accumulation
of contaminated sludge sediments  for about 6 years.9 The PCB
concentration in the sediments averages  100 to 300 ppm.

PRACTICES TO CONTAMINATE SEDIMENTS
  Much of the contamination in the water sediments is the result
of direct or indirect discharges from the industrial manufacturing
processes. The major users of PCBs, a major component of the
contaminants, are electrical capacitor manufacturers. Although the
use of PCBs is banned by the U.S. EPA at present, active discharges
of PCBs to the waterways along with localized PCB spills  when
regulatory standards  were lacking  were the major causes of the
sediment contamination.
  The PCB wastes discharged to the nearby municipal sewage sys-
tem also resulted in sludge contamination. In addition to the release
of contaminants from  sludge disposal, the absence of the secondary
treatment step common in the treatment plants located along the
coastline resulted in a pervasive pollution problem throughout the
widespread areas where effluent discharges occurred. Since PCBs
are highly resistant to biological decomposition, the biological treat-
ment step probably would have had the minimal effect of reducing
PCBs or other pollutant concentrations in the discharge streams.
  It also has been recognized that the long-range transport of PCBs
or other toxic organics with low vapor pressures which are persis-
tent in the atmosphere can eventually precipitate and enter the water
resulting in contaminated sediments.  Good examples  of this
phenomenon are the presence of trace amounts of PCBs in the polar
capice and the ubiquitous presence of PCBs in the sediments of
the Great Lakes in trace quantities. These levels of contamination
will represent background concentrations and will not present  a
controlling influence on the result of the public health assessment
for contaminated Superfund sites or on the evaluation of remedial
486    MULTI-MEDIA

-------
action alternatives concerned with the removal of the bottom sedi-
ments with high levels of contamination.

CONTAMINANT MIGRATION
  Physical transport processes have moved the contaminants and
the sediments slowly downstream of the discharge locations. Sus-
pensions being discharged would have slowly settled to the bottom
as they were carried along with the movement of the water.  The
sampling data show that the range in contaminant concentrations
in the waterway sediments is very wide. In addition to the wide
ranging spatial variation of the contaminant concentrations, the
sediment PCB data also show the vertical change in the concen-
tration distribution. Hot spots representing areas with the high con-
centrations of contaminants are mostly at the upstream of the
flowing water. Contaminant concentrations generally decrease
farther away from the discharge points. Floods and storm events
are known to have an effect on the original distribution patterns.
  When enough data are available, delineation of concentration
distribution in the form of a color coded map or contour lines will
aid visual observations of the migration of contaminants. However,
it should be noted that the contaminant distribution is a dynamic
and ever changing process. Oftentimes, it is difficult to identify
the physical processes responsible for the transport and disturbance
of the bottom sediments. The upstream mud flat areas would still
contain elevated concentrations of contaminants, probably because
the sediments arc flushed less rapidly. However, both the aerobic
and anaerobic portions of the  mud flats would be subject to
photolytic decomposition and biodegradation more frequently than
those sediments in the water column. The  apparent  degradation
of low molecular components of PCBs has been reported although
it is not yet clearly  defined.
  Several investigators have proposed models to describe the trans-
port of sediments in  the harbor or estuary. The models incorporate
storm and runoff events, scouring phenomena,  settling, suspen-
sion, sedimentation and partition coefficient for adsorption and
calculate the concentration variation as a function of spatial coor-
dinates.8 Resuspension occurs  via the  shearing action of the
flowing water and through the actions of organisms that inhabit
the upper layer of the sediments. Kinetic factors that influence the
level of contaminant concentration in sediment include  biode-
gradation and volatilization.  Storms and other events causing dis-
turbances in the water column result in increased resuspension of
sediments. The amount of contaminants which will be further
carried downstream  due to the storm event will be dependent upon
the settling velocity of the suspended particles and the distance that
the particles travel.  Additionally, the transport of sediments will
be affected by the frequency of storm events, volume of runoff
and rainfall intensity.
  Mud flats with vegetative cover tend to be less subject to scouring
than the main river  body because the vegetation tends to immo-
bilize sediments in the area. Also, ice that forms during  the winter
months may be an important mechanism for transporting the con-
taminated sediments from wetlands. The mats of mud  flats sedi-
ments that are frozen underneath the ice can be carried away when
water levels  rise and water flows during the spring thaw.

EXPOSED ECOLOGICAL  SPECIES
  The various species of life must be characterized to  assess the
impact of the contaminants on the ecosystem. Since it may be im-
practicable to fully characterize a variety of species in the affected
area, it is convenient to identify typical or  most affected subsets
of the species classified according to the ecological system where
the species inhabit or according to the medium pertinent to the
life cycle of the  organisms.
  The first type  of classification can be divided into the  aquatic
species and the terrestrial species. The aquatic species are  comprised
of flora including aquatic vegetation which provides cover for fauna
and food for waterfowls. Wetlands, which support marsh vegeta-
tion and are breeding and  nesting grounds for wildlife. Also
 included here are fauna including zooplankton, snails and a variety
 of fish and shellfish.
   The second major classification considers the medium through
 which human exposure occurs, such as sediments, water column
 and food webs. Then the system identifies various species that are
 affected by the contaminants under that particular subclassifi-
 cation. According to this classification, the fish species can fall
 into each of  the three subclasses:  species inhabiting the water
 column; species inhabiting the sediments such as some member of
 mollusks and arthropods; and species important in food chain webs
 which also include mollusks, arthropods, zooplankton and certain
 fish.
   Organisms  in a community, plus environmental  factors with
 which they  interact, comprise an ecosystem. To have balanced
 ecosystems, the ecological risks associated with direct and indirect
 pathways affecting  aquatic and  terrestrial lives should  not  be
 adverse. Since  there are many  factors  affecting relationships
 between organisms and between organisms and the environment
 and since these  relationships can be very complicated to under-
 stand, species affecting all the basic  components of the ecological
 relationship should be identified. Basic aspects of the relationships
 in evaluating ecological species include: impacts on producers, con-
 sumers and decomposers; flow of energy in the food web cycle and
 representation of trophic levels; affect of abiotic limiting factors
 on the organisms affected by the contaminants; and a possible
 breakdown  of the fixed relationship between organisms.
   In addition to primary impacts on the exposed ecological species,
 secondary impacts resulting from no action or removal of sedi-
 ment deposits can affect the balance  of the ecological species. The
 secondary impacts arise from the continued closure  of fisheries,
 the prohibition of maintenance of dredged material and the con-
 tamination of downstream water supplies.

 PUBLIC HEALTH EXPOSURE ASSESSMENT
   Exposure pathways that lead to adverse health effects resulting
 from exposure to contaminants directly or indirectly released from
 the sediments into the environment should be identified as a first
 step in the human health risk assessment. Once all the routes of
 exposure contributing to the overall health risk are identified, a
 subdivision into significant and insignificant routes could simpli-
 fy the subsequent effort of analyzing and collecting data to evalu-
 ate the phases of exposure assessment and risk analysis.
   Subsequent analysis involves gathering data or estimating ranges
 of values concerning  routes,  frequency,  duration  of human
 exposure for each pathway, the size and activity patterns of the
 population affected, environmental concentrations of contaminants
 and exposure levels for major pollutants of concern to which popu-
 lations may be exposed. The analysis should also consider conse-
 quences of exposure resulting from  remedial alternatives such as
 the removal of contaminants from the waterways. Public health
 impacts which include long-term  and short-term effects are part
 of primary impacts. The long-term health effects result from long-
 term exposure to the contaminants possibly over a person's life-
 time and often use the terms carcinogenic and non-carcinogenic
 effects. The short-term health effects can occur upon exposure for
 a short-term period lasting for  1 day, 10 days or longer but less
 than a  lifetime period.
   Table 1 shows an outline of major exposure pathways pertinent
 to the evaluation of exposure from  contaminated sediments and
 some ecological species  that can be exposed to sediment con-
 taminants. Exposure pathways  are  divided into two scenarios
 representing no  action alternatives and the short-term and long-
 term impacts associated  with alternatives of  possible removal
 actions.
   Depending upon the expected magnitude of human exposure the
 range of exposure estimates can be  performed: typical case esti-
 mates and reasonable  worst case  estimates.  The  typical case
 represents the best estimate of a mid-range value for parameters
.influencing the likely exposures. The reasonable worst case refers
                                                                                                      MULTI-MEDIA    487

-------
to the most conservative expectation for the parameter value and
occurrences which may represent the upper-bound estimate  of
human exposure but can be expected in the worst case situations
without being unrealistic. Although the exposure estimation pro-
cedure described here can be extended to the reasonable worst case
scenario, it is more oriented toward  assumptions which tend  to
represent the typical case scenario.
                           Table 1
     Exposure Pathways Associated with Contaminated Sediments
                 And Major Cleanup Alternatives
 No Action
 InhaLulion from  wolatIlizalion &
           paniculate suspension
 Ingestion :  fish
             water
             sediments
             hunted wildlife
 Dermal Absorption  : water
                    aediments
Romcri>aLIon  Act Ion
o Dredging component
   Inhalat ion
   Dermal  abaorptIon:water
o Diapoaal Component
   Inhalat ion
   Ingeation  :  ground water
               aol 1
               vegetat Ion
                           due to Contaminated Sediments
   posed Ecological Specie
      Aquatic
         Flora
         Uetlands  ( habitats for wildlife  )
         Fauna
         Fish
      Terrestrial
         Flora
         Fauna
         Wildlife
         Agriculture
      Endangered species
      En v i romen t a 1 1 y receptive areas
Air Inhalation
  A contaminant may volatilize from the water column after trans-
port from sediments through diffusive or bulk motion phenome-
na; or it may evaporate directly from the sediment surface when
it is exposed to the air. In the former case,the whole water body
overlying the contaminated sediments can be a source of toxic air
emissions. The direct emissions from the exposed surface can oc-
cur from the mud flats area and when the levels of water are receded
because of low tide or dry season. Hence, care must be taken in
defining the background air concentrations from ambient air
monitoring data. The true background should represent the am-
bient  air concentrations not affected by the emissions from the
water column.  Since this emission label is comparatively lower than
that from exposed sediments or, from sediment undergoing dis-
turbances which  can occur during the dredging operation or other
activities, the baseline background attributable to the ambient air
from  the water column  may be difficult to distinguish from the
true background concentration.
  Ill-defined monitoring data can be confusing in exposure assess-
ment. Because the exposure concentrations are dependent upon
many factors including emission rates influenced by concentrations
in sediments, tidal cycles, distance to the affected population and
wind-rose characteristics,  it  is  often  convenient to apply air
emissions and  dispersion models  to assess the potential of human
health risks.  Air emission models  for high  tide, low tide and
exposure sediments are  presented below.

High  Tide
  The high tide situation represents volatilization occurring from
the water column with a relatively nonflowing water and a reason-
able depth of  water above the contaminated sediments. The rate
of volatilization  will be affected by various chemical and physical
parameters which are site—specific and chemical specific. Some
of these parameters affecting the transport rate are vapor pressure,
Henry's law constant, diffusion coefficients, thickness of diffu-
sional resistance in sediment and mass  transfer  coefficients. The
two-resistance theory under the assumption of a steady-state
                                emission rate and simplifications of the resulting equation provide
                                the following formula for the emission rate:
where q  = the flux rate, g/cm2 .sec; k =  sediment-side

mass transfer coefficient, cme/sec;  Kd =  partition

coefficient between sediment and water, L/Kg; and Ce = con-
taminant concentration in sediment, g/g.  Because of the length
involved in the derivation of this simplified equation, its detail
derivation is not provided here. In Equation (1), the sediment side
mass transfer coefficient can estimated from
k  =
            /r
                     cm/sec
(2)
with Dw  = diffusion coefficient of contaminant in water,
cm2/s, E  = porosity of the sediment  and r  = diffusion path
length below the water/ sediment interface, cm. (Use about 5-10 cm
for default sediment thickness).

with D = diffusion coefficient of contaminant in water, en 2/s,
E  = porosity of the sediment and r ' diffusion path length below
the water/ sediment interface, cm. (Use about 5-10 cm for default
sediment thickness).

Low Tide
  The relatively high velocity created by the receding water flow
during low tide results in increases in the exchange rate of the con-
taminant  between the  sediment and water  phases and a slight
resuspension of the sediment layer, thus increasing the mass transfer
rate. The application of the two-resistance  theory provides the
following equation for estimating emission  rates.
                                q  =
                                       ke kw KซL
                    C,
                                                                                         (3)
                                     (KoLKd +
                                Where kป =  water side mass transfer coefficient at the velocity
                                of the low-tide  water, cm/s,  and K^  = overall mass transfer
                                coefficient for the air-water interface, cm/s.

                                Exposed Sediments
                                  Once the overlying water has receded, the resistance to volatili-
                                zation is decreased by the amount created by the overlying water.
                                At the early stage of the volatilization process, the pore space within
                                the sediment particles could be filled with water in the absence of
                                good drainage. In this case, the diffusion should occur  along the
                                water in the pores. In the presence of good drainage or after a long
                                time of exposure of the sediments in the atmosphere, the pore space
                                can become unsaturated. In this  case, the diffusion should occur
                                along the path of air pores until  the molecules reach the sedi-
                                ment/air interface for dispersion.
                                  The steady-state emission rate as well as transient emission rate
                                in which  the  emission rate changes over time after the  initial
                                exposure of mud flats in contact with the air can be independently
                                derived. The results for emission rate from unsaturated sediments
                                in contact with air can be  presented as  follows:
                                • Transient emission rate averaged over an emission period.
                                     2EDC,
                                      irv T
                                                         (4)
                                where   D,.,  = effective   diffusivity   (DiE1"),    cmVsec;
                                D| =  molecular diffusivity in air, cmVsec; E  = fraction of un-
                                saturated sediment filled with air, cm' air/cm^; H = Henry's law
                                constant, cmJ H2O/cm3 air ( = 41 atm mVg mol); T  = exposure
                                period  over which  emissions are averaged,  sec; *  = D
488     MULTI-MEDIA

-------
E/(E + QS (1  - E) Kd/H); gs = true density of sediment, g/cm3;
and the symbol - over q represents the average of q.

•  Steady-state emission rate.
   The steady-state emission rate from the unsaturated sediment
can be derived from the two-resistance theory.
        ฃ3/2   CeH
                Kd
                                                         (5)
 where r = average diffusion path length for a contaminant in sedi-
 ment, cm and other symbols are as defined previously. The value
 for E can range from 0.1 to, 0.4, the representative value being
 about 0.1 to 0.2. The emission rate equations similar to Equations
 (4) and (5) can be derived for the case where the sediment in con-
 tact with the air is saturated with water. For the sediment saturat-
 ed with water, the concentration in the water phase constitutes the
 driving force for diffusion as compared to the concentration in
 the air pores in the case of unsaturated sediments. Hence,  the air
 phase diffusivity in Equations (4) and (5) should be replaced by
 the water phase diffusivity D2 and Henry's law  constant is not
 needed since the concentration distribution between the water and
 sediment phases is related by the  partition coefficient. The term
 Kd/H in the two equations should also  be replaced by Kd.  The
 porosity term E should represent the saturated porosity and may
 range from 0.3  to 0.9 depending upon sediment type.  Porosity
 values for clay sediment may be in the range of 0.5 to 0.6.
   Validation of the air emission models is not complete because
 of a lack of  monitoring data  obtained under  uncontrolled
 experimental conditions identifying  all  the necessary variables
 needed in the validation. A sample calculation predicted PCB con-
 centrations in the air above a harbor water body in the neighbor-
 hood of 5 to 10 ng/m3. These values compared  fairly well with
 the measured ambient air quality. Efforts are  underway to inter-
 pret the data more extensively and to gather additional data useful
 for low tide and mud flat  emission scenarios as well as  high tide
 situations.
   The estimated emission  rate can be substituted into dispersion
 equation to predict the ambient air concentrations at a receptor
 location above the water surface  or beyond the downwind edge
 of the water body. Different dispersion formulas apply for the two
 situations. These dispersion equations for calculating ambient air
 concentrations  of contaminants follow.
 • Above the water surface
 C =
            -    [
            o,—
= erf (
                   (6)
 where C = ambient air concentration, g/cm3; x = distance from
 the  upwind edge  of  the  contaminated  water  body,  cm;
 5Z =  standard deviation in the vertical direction, cm; u = wind
 speed, cm/sec;  and z = the receptor height, cm.

 •  Beyond the downwind edge of the contaminated water body:
   The ambient concentration at a receptor location beyond the
 source boundary can not be estimated by Equation  (6) which is
 designed for on-site concentrations. The ambient air concentra-
 tion off-site of the water body or exposed sediments can be esti-
 mated by using the following equation:
C  =
     2 TTU 5Z
  erf
_erf(y-b/2.
  **1 * \     f.
                                                         (7)
where a  = equivalent length of the contaminated water body
parallel to the wind direction, cm; b = equivalent length of the
dontaminated water body perpendicular to the wind direction, cm;
                                              y = distance measured perpendicular to the wind direction from
                                              the center line of the winds, cm; and 6j  = standard deviation in
                                              the lateral direction, cm.
                                                The following relationships for estimating lateral and vertical
                                              standard deviations should be used:
                                                •  For stability C
                                                               0.894                                  (8)
                                                    y = 0.35  X
                                                               0.911                                  (9)
                                                    z = 0.17  x
                                                • For stability D
                                              where x and y are in cm.
                                                               0.894
                                                    y = 0.23  X
                                                               0.725
                                                    z = 0.78  x
                                                                                                                         (10)

                                                                                                                         (U)
  The approach employed in Equations (6) and (7) assumes that
the emission rate over the entire area of contamination is uniform.
This situation may rarely occur at real sites because of a wide
ranging spatial distribution of contaminants concentrations in sedi-
ments. This problem can be remedied by dividing the area of emis-
sions into smaller sections where each section can be a source of
a uniform emission. Equations (6) or (7) or the Industrial Source
Complex (ISC) models can  be  applied to emissions from each
section.
  The ISC models assume a virtual point source for emissions of
an  area source and calculate the downwind concentrations at
ground level using steady-state Gaussian plume models. The con-
centrations at a population receptor will be a summation of the
all the concentration contributions from the sectionalized emission
sources.
  The atmospheric concentrations are used to estimate short-term
or long-term exposure to contaminants originated from sediments.
Standard exposure parameters needed in the estimation include the
breathing rate of the exposed population, frequency and duration
of exposure, information regarding the extent of absorption of the
inhaled air contaminants and the body weight distribution of the
affected population segment.

Dermal Contact
  Exposure  through dermal contact can occur  from  several
activities which result in contact with water or sediments. Dermal
contact can arise from washing sediments off a boat,  swimming,
fishing, wading and boating.  The importance of exposure during
these activities can vary depending upon activity patterns of an
individual. For boat keepers, sailors and fishermen, it is likely that
activities relating to boating and fishing will result in the highest
exposure, since they may spend more time hi these activities. People
may stand in the mud flat area during fishing, resulting in dermal
contact of the lower body part with the sediment. Swimming will
lead to whole body contact with the swimming water and hence
dermal absorption of the contaminants through the  skin.
  The exposure  estimation  for the dermal contact pathways
requires information regarding the time spent for each activity and
its frequency, surface area of the skin exposed to the water or sedi-
ments, absorption rates which are dependent upon the types of
contaminants and the medium  on which the  contaminants are
absorbed, contaminant concentrations and the individual's body
weight.
  The typical body weight in the exposure calculation is 70 kg and
the exposed skin area  of the typical  human body is 18,150 cm2.
When the low parts of both  leg areas and hands are submerged
into the water during fishing, exposed areas would approximate
2,800 cm2. Presently, very little information is available for ab-
sorption of sediment contaminant through the human skin.  For
PCBs absorbed on soil, U.S.  EPA reported an absorption rate of
                                                                                                      MULTI-MEDIA     489

-------
5%. It is possible that the higher organic content in sediments than
in soil may reduce the absorption rate for the contaminants in sedi-
ments. For the comparable organic contents (1 to 5 % range), it
may be reasonable to assume a 5 % absorption rate for PCBs in
sediments.

Fish Ingestion
   Fish are exposed to contaminants through contact with water
and sediments and by preying on smaller organisms inhabiting the
contaminated area. The concentration in fish is usually higher than
the contacting medium because the lipid portion of fish can easily
bioconcentrate the organic compounds. This is especially true for
PCBs, the major components of lake or harbor contamination.
Fish as used here refer to aquatic biota that includes mollusks and
arthropods. Fish caught commercially and as a sporting event from
the contaminated area  should  be accounted for  in the exposure
analysis.
   The major parameters affecting the results of exposure for the
fish ingestion pathway include the fish ingestion rates for each type
of fish living in the bulk of the water column and in the sediments,
the number of fish meals eaten by the exposed person,  fraction
of fish that comes from the contaminated area and the contaminant
concentration in fish. The distinction  between fish living in the
water and sediments can be important in estimating the con-
taminant concentrations in fish, because the contaminant uptake
rate in the sediment-loving fish is more affected by the concentra-
tion in sediment than in water.  For these species,  it is appropriate
to use the fish to sediment bioconcentration factor to estimate the
extent of contaminant bioaccumulation in fish. The species of fish
living in or  close to sediments include flounders, eels, lobsters,
crabs, mussels, clams and snails.  The estimated concentrations
should be compared to monitored fish contaminant levels to define
the range of contamination in various fish types.
   Although the fish consumption is somewhat contradictory in
various  literature, the annual average consumption rate  of fresh
fish ranges from 6.3 to 10 g/day. The average consumption rate
for a fish meal is much higher, amounting to about 225 g/day.
Heavy fish eaters and fisherman can consume fish substantially
higher than  these average values in quantity.

Water Ingestion
  Exposure  to contaminants released to the water column from
the sediments could occur  due to ingestion of  water during
swimming. The parameters affecting the magnitude of exposure
include the amount of water ingested, contaminant concentration
in water, frequency and duration of swimming and body weight.
Typical  values for the swimming events can range from  20 to 30
times a  year, ingesting  the water at the rate of about SO ml/hr.

INTEGRATED EXPOSURE ANALYSIS
  The integrated exposure analysis combines the exposures esti-
mated for each pathway with toxicity information for each con-
taminant of major concern found in the sediments. This step
provides a  quantitative  estimation of  the  public  health risk
associated with the release of contaminants from the sediments.
The public health risks can be associated with short-term and long-
term levels of contaminant exposure or the short-term or long-term
affects of exposure.
   Multimedia risk can be obtained by summing the risks over the
total number of exposure pathways considered. It is in accordance
with the U.S. EPA guidance to add the ratio of the dose level to
the acceptable dose  level for  each contaminant over the total
number of pathways in dealing with non-carcinogenic risk evalu-
ations.  If the summation of the ratios for the non-carcinogenic
effects exceeds unity, the contaminants in sediments pose an un-
acceptable level of non-carcinogenic risk. The risk estimation can
also be combined with information concerning the number  of popu-
lation exposed to the contaminants. The result is an estimation of
the number of population which may be affected by the long-term

490    MULTI-MEDIA
or short-term effect.

REMEDIATION OPTIONS
  There are a number of alternatives being evaluated to alleviate
or eliminate the public health risks associated with contaminated
sediments. If remediation options are considered, the health risk
assessment  should address the public health and environmental
impacts of the feasible alternatives in terms of long-term and short-
term consequences and the impacts associated with the disposal
or treatment of the  dredged material.
  Major types of feasible alternatives frequently discussed include
insitu containment and clamshell and hydraulk dredging. Insitu
containment involves isolation of hot spots using dikes and benns
that are built in  waters of suitable depth.  An added precaution
includes use of a clay cap placed over the sediments or a plastic
liner plus silt and rocks. The potential problem with insitu con-
tainment arises from potential navigational hazards and difficulties
with the construction of containment structure. Also, when using
insitu containment,  one should stabilize  hot spots. Plastic liners
can be ruptured and the fine material underneath the liners could
be scoured  during high flows.
  Clamshell dredging involves clamshell buckets mounted on barge
cranes. The excavated material is transported to a handling area
and for subsequent  transport to a containment disposal facility
(CDF)- The quantity of material being excavated will be limited
by the speed that the buckets can pick up sediments. The hydraulic
dredging equipment sucks the sediment material through the suction
dredge which is in contact with the sediments. A hydraulic pump
provides power for suction of the sediments which can be hydrau-
lically transported to the nearby CDF because sediment  slurries
are handled in pipes. There are other dredging systems available.
However, it appears difficult  to justify these other systems over
the clamshell or hydraulic dredging system at this time.
   In  all of the  remedial  alternatives considered, disturbances
occurring during the dredging or other  remedial operation will
resuspend the  contaminant particles. There is the likelihood of
increased volatilization or suspension of fine particles in the atmos-
phere, causing short-term increases in the contaminant levels. Air
monitoring as well as monitoring for other media must be con-
ducted carefully during experimental dredging operations to dearly
define the levels of hazardous contaminants in the ambient air in
relation to the affected population. An adequate, prior assessment
for potential health  risk is important when the full-scale remedial
action is contemplated. Models are extremely convenient in this
respect because once the emission rate at the dredging source is
estimated or calibrated through monitoring, the ambient air con-
centration at the population location can be easily estimated.
   Temporary storage and ultimate disposal of dredged material
should also be evaluated for potential public health  risks. If the
ultimate disposal site considered is near a body of water, long-term
erosion of  the waste material may again contaminate the water
body. The  potential impacts of air emission and leachate migra-
tion on public health risk can be a primary concern  of exposure
assessment. The potential  for public health risks associated with
the disposal and storage of the excavated sediments cannot be sepa-
rated from the integrated risk assessment. If off-site transport of
excavated sediments is evaluated,  accidental spills during transpor-
tation are also a possibility.

CONCLUSION
   This paper identifies exposure pathways pertinent to evaluating
the potential public health and ecological risks associated with con-
taminated sediments in a marine environment. Exposure assess-
ment methodologies also are described. Models useful in estimating
volatilization rates from contaminated areas are presented for cases
corresponding to high tide, low tide and mud flats and should help
evaluate the population exposure affected by inhalation during
implementation  of no action  or  remedial action alternatives.

-------
REFERENCES
1.  Bopp, R.F., "The Geochemistry of Polychlorinated biphenyls
   in the Hudson River." Ph.D. Dissertation, Columbia Univer-
   sity, New York, NY 1979.
2.  Brown, M.P., Werner, M.B., Sloan, R.J., "Polychlorinated
   biphenyls in the Hudson River," Environ. Sci. Technol., 19,
   1985, 656-661.
3.  Hwang, S.T., "Methods for Estimated On-Site Ambient Air
   Concentrations at Disposal Sites,'' Nuclear and Chem. Waste,
   Manag., 7, 1987,  1-4.
4.  Malcolm Pirnie, Inc. "PCB hot spot dredging program, Upper
   Hudson River, New York." Draft environmental impact state-
   ment. Prepared for New York State Department of Environ-
   mental Conservation, Albany, NY, 1980.
5.  U.S. EPA, "Acushnet Estuary PCBs Data Management Final
   Report, Region 1, Contract No.68-04-1009," Aug. 1983.
6.  U.S. EPA, "Industrial Source Complex (ISC) Dispersion Model
   User's Guide. Vol. 1. EPA-450/4-79-030," Dec.  1979.
7.  U.S. EPA, "The PCB Contamination Problem in Waukegan,
   111.," Region V, Jan. 1981.
8.  U.S. EPA, "Mathematical Modeling Estimate of Environmental
   Exposure due  to  PCB-Contaminated Harbor  Sediments of
   Waukegan Harbor and North Ditch, Cincinnati, Ohio." Feb.
   1981.
9.  Weaver, G., PCB contamination in and around New Bedford,
   Mass., Environ. Sci.  Technol., 18, No.  1, 1984, 22A-27A
                                                                                                 MULTI-MEDIA    491

-------
                          Earned  Value as  an  Appropriate  Tool
                    For Controlling Remedial Planning  Contracts
                                     u.s
    James B.  Chaffee, Jr.
      Robert T. Fellman
Ebasco Services Incorporated
      Arlington, Virginia
         John J. Smith
 Environmental Protection Agency
       Washington,  DC
ABSTRACT
  Baseline management with Earned Value is a project control
methodology employed by many contractors in private industry
and is a requirement on government contracts with agencies such
as the Department of Energy and the Department of Defense.
However, in the environment of remedial planning for hazardous
waste sites, does this methodology provide an effective and reason-
able  means to measure and control Program  work?
  This paper defines the Earned Value concept and analyzes its
usefulness in controlling remedial planning contracts.  The interpre-
tation of the Earned Value calculations is discussed as to their
proper evaluation.


INTRODUCTION
  As the volume of work  under  the Superfund program has
expanded,  and the number of active sites and contractors  per-
forming work at these sites has multiplied, the need for a uniform
method of cost/schedule control of these efforts has increased.
There are  many  different approaches  to controlling  projects,
however; the Earned Value approach to program control offers
the most flexibility because it allows both  in-depth review and
"analysis at a glance" of diverse program elements.
  The purpose of this paper is to define the meaning of the Earned
Value concept and show how it is used in program  control.  The
focus of this paper is on the type of assignment likely to be received
under a U.S. EPA Superfund contract, i.e., Rl/FS. Earned Value
as a  tool for  project  control however is relevant  and applicable
to virtually any type of tasked, plannable effort.
  Before proceeding  further, it is useful to establish the project
control requirements  are for a task order-type contract. We must
first state that we  believe task order-type contracts require a level
of program management equal to, if not more rigorous than, any
other form of contract.
  Most approaches to  program control are extreme.  The  first
extreme is  to do too little. This approach says that rigorous  pro-
gram controls cost too much money, and eliminating them saves
money. The major argument used to support this view is that a
technically  knowledgeable work assignment team led by a conscien-
tious site manager can concentrate on the work  assignment deliver -
ables without having to dilute resources for project controls which
contribute nothing to the end product.
  The second extreme is to do too much. In this approach, overly
detailed and intricate CPM schedules are prepared. Budgeting is
done at a very detailed level. The major argument  used for this
approach is that a detailed definition of the work scope prevents

492    COST AND  ECONOMICS
                    later disagreements about what work is in scope and what work
                    is out of scope.
                     As with all extremes, the best approach is somewhere in between.
                    The need to develop a work plan which can be monitored and con-
                    trolled at a level of detail commensurate with the total value of
                    the work assignment  is clear and absolute.  Using this approach,
                    the cost of controlling work assignments can be kept in the 2 to
                    3% range.

                    METHODOLOGY FOR PROJECT CONTROL
                    Definitions
                     There are five basic data elements used in an Earned Value
                    System. These data elements are the basis of variance calculations
                    which yield performance information necessary for project control.
                     The following is a description of these key Earned Value data
                    elements:
WORK TASKS
NO
I
3
4
BUDGET
W
ISO
too
SO
MONTH
MM | fee | MAP | APR | UAT | JUNE
4 *
f BUDGET JO J
[ }- EAflXD UU.UC W
| *TT IMI 3D |
A &
[ BUOGCT 190 J
[ 	 J- EARNED VULUf n
( ACTUAL 130 J
A A
| BUDGET K 20 |
[ EARNED VM.UE W 1
[ ACTUAL W
A 4
| BUDGET SO j

                                                                   A SCHEDULED
                                                                   AACTUAI
                                      APP. ปh BUDGET
                                      APH Mm EARNED VW.UE
                                      APR 30th ACTUAL COST
                                      BUDGET AT COMPLETION
                                      ESTIMATE AT COMPLETION
                                      H COMPLETE
                                      COST VARIANCE
                                      SCHEDULE VARIANCE
mo
715
Jซ0
ISO
310
61 
-------
  Budget—Budgeted Cost of Work Scheduled (BCWS). BCWS
is the budget value of the work scheduled to be accomplished within
a given time frame (Figure 1). In other words, this is the budget
value of the work which was supposed to be done by a given status
date.

Earned  Value—Budgeted Cost of  Work Performed (BCWP).
BCWP is the budget value of the work actually accomplished in
a given time-frame (Figure 1).

Actuals—Actual Cost of Work Performed (ACWP). ACWP is the
cost for completing the work actually accomplished. This should
be expressed both in terms of workhours and dollars (Figure 1.)

Budget At Completion (BAC). BAC is the total budget established
and approved for the completion of a work assignment.
Estimate At Completion (EAC). EAC is the sum of the actual cost
plus the forecast to go.

  Various  calculations using these five basic data elements yield
performance indicators useful in determining  work assignment
status and  potential problems. We will now discuss the main cal-
culated performance indicators.

Percent complete—This is  the relationship between the Earned
Value and the Budget  at completion.
                              BCWP
               % Complete =
x  100
(1)
   In other words, if the budget value of work actually accomplished
 is 60% of the total budget for a work assignment, then the work
 assignment is 60% complete.
 Cost Variance—This variance is a measure of the cost overrun or
 underrun of the budget established for the work accomplished to
 date based on the earning rules in effect and the actual cost to date.
 Cost Variance =BCWP - ACWP
                    (2)
 Schedule Variance—This variance is the difference between the
 Earned Value and the Budget value of the Work scheduled. The
 Schedule Variance calculation is a quantitative measure of schedule
 slippage which is calculated as follows:

 Schedule Variance  = BCWP - BCWS                    (3)

   These calculations yield performance indicators, both for cost
 and schedule. These indicators are not in any sense absolute, except
 at the completion of a work assignment. At intermediate points
 they indicate potential problems which should be investigated to
 determine if they are real or transient. If they are real, the impact
 on the work assignment can be evaluated and decisions made as
 to the appropriate course of action to be taken early in the work
 assignment. If they are transient,  no action is necessary.
   One of the most  useful aspects of utilizing the Earned Value
 approach to project control is in  the trend analysis of the data
 elements. This may indicate an impending problem long before
 the problem would be reflected by comparing invoices to funding
 limits.

 HOW DO YOU CALCULATE EARNED VALUE?
 Work Assignment Baseline
   To monitor and control program schedules and budgets, a base-
 line must be generated for each work assignment. It is composed
 of three elements. The first is the technical baseline which defines
 what work will be done, how it will be done and what deliverables
 will be produced. The second baseline is the schedule, which defines
 the  activities and their  sequence in  time  for  doing  the  work
 according to the technical baseline. The third baseline is the budget,
 which defines the cost (both labor and non-labor) for performing
 the  work according to the technical and schedule baselines.
   The first consideration in baseline development is to assure tech-
 nical understanding of the scope of the assignment. One process
which the authors have found helpful in scoping the RI/FS is the
use of the DQO Process, which has emerged via considerable U.S.
EPA guidance. The DQO Process involves developing a prelimi-
nary risk assessment, identifying of remedial technologies,  early
assembling of the ARARs and understanding their applicability
and, finally, specifying field techniques and laboratory analyses
appropriate to the level of data quality needed to support decision-
making. The technical baseline is developed when the team can
produce a task-activity network showing the interrelationships in
logic.
  Identifying the required technical activities and arranging the
logic of their interrelation is a necessary condition to proceeding
to the calculation of Earned Value.  The technical baseline  is
brought to the next level of organization by:

• Classifying the activities  into recognizable tasks
• Specifying the duration of each activity

  It is very useful to annotate  for each activity any important
assumptions guiding one,s expectation of its duration; for example,
it may be assumed that two 30-ft wells per day, per rig is achieva-
ble at the site, or that the U.S. EPA's obtaining access to the site
will require a minimum of two weeks given the need to involve
the Office  of Regional Counsel. Having identified durations for
each activity, an overall schedule can be developed. A critical path,
the sequence of logically connected activities which will yield the
longest path in time from start to finish for a given (total) action,
can be determined easily from this schedule.
   For the final step, preparing the budget baseline, each activity
is priced. An example of a tool which is very helpful in developing
accurate budgets is a priced generic schedule. All major unit oper-
ations or activities are incorporated into the generic schedule and
best estimate pricing is  provided both for labor as well as  other
direct costs. The estimates in the priced generic schedule can then
be compared to the assignment specific cost estimates on an activity-
by,  activity  basis.  The cost estimate  presented in the generic
schedule then is adjusted in the site-specific estimate to account
for the scope of  the assignment.
   At this  point,  all information,  (technical, time  schedule and
budget information) is available to compute the budget baseline.
Pictorially, the basic process of assembling the baseline is shown
in Figure 2 where the budget baseline is presented as a function
 of time. This functional representation is the Budgeted Cost  of
Work Scheduled (BCWS) previously defined.
                                1  DEFINE THE WORK
                                2. SCHEDULE
                                  THE WORK
                                                          Figure 2
                                                  Work Assignment Baseline

                                                         COST AND ECONOMICS     493

-------
EARNING RULES
  Once the work assignment baseline is established, the earning
rules determine the time-phased budget and Earned Value. The
two most widely used earning rules are the 50/50 rule and the linear-
by-duration rule (Figure 3). The use of pre-defined earning rules
eliminates the inherent subjectivity of a  human evaluation  of
progress. These  earning rules are simple and are applied at the
activity level.
  The 50/50 rule specifies that 50% of the activity budget will be
earned the day the activity is started. Nothing further is earned
until the activity is complete, at which time the  remaining 50%
is earned. This  earning rule is  appropriate for  short duration
activities with relatively small budgets (e.g., activities with a
duration of less than 8 weeks). If an activity duration is short but
the budget value is high (e.g., site investigation activity), the 50/50
rule should not be used. In such cases the linear-by-duration rule
is preferred.
  The linear by duration rule generally is used for an activity which
lasts longer than 8 weeks. The Earned Value is  the ratio of the
activity duration to the status date to the total duration times the
total activity budget.
  By utilizing objective earning rules, a standard approach to the
evaluation of project/work assignment progress is adopted so that
when  a work assignment is reported to be 60"% complete,  its
meaning can be understood and appreciated.
  For  the calculation  of budget (BCWS), the controlling factors
are the scheduled start and completion dates and activity duration
in the  target  (baseline) schedule. For the calculation of Earned
Value (BCWP), the controlling factors are the actual start and com-
pletion dates and  the activity duration in the current  (statused)
schedule. The statused schedule may be a shortened or lengthened
form of the target schedule depending upon project exigencies and
the resources brought to bear on the activity. The statused schedule
is dynamic, thus the rate of earning can be increased or  slowed.

EARNED VALUE ANALYSIS
  Using the Earned Value Data Elements, the status of each work
assignment and/or higher level of summarization can be readily
determined. For example, Graph A on  Figure 4 shows baseline
budget and actual curves for a hypothetical example built around
a hypothetical work assignment. The actual cost is less than that
budgeted, which is good, but there is no indication as to whether
the scheduled work has been accomplished.
  Graph B adds the Earned  Value data. In this case, it has been
assumed that the Earned Value curve lies above both the budget
and actual curves. This placement signifies that the Work Assign-
ment is both ahead of schedule and under budget. The project is
ahead  of schedule since the  Earned Value,  i.e., the accomplish-
ments  exceed the  budget. Since the Earned Value exceeds  the
actuals, the project is also under budget.
  If the Earned  Value curve lies between the budget and actual
curves (Graph C), the work assignment is behind schedule but under
budget. If it lies below  both curves (Graph D), the work assignment
is both behind schedule and over budget.
  The  addition of Earned Value data allows a manager to scan
quickly work assignments to  determine their status. For example,
to determine the percent complete, the Earned Value to that point
in time is divided by the total budget  (i.e.,  BAG) and expressed
as a percentage. Analysis of Earned Value also allows for the ear-
ly warning of undesirable trends and provides for timely correc-
tive action.
  The  ability to do trend analysis is an important result of using
the Earned Value approach to program control. By examining the
Earned Value curves,  it is possible to identify,  based on  the rate
of expenditure, an impending cost overrun. This can be done prior
to actually overunning the costs on the assignment. Additionally,
by examining the rate of earning, a potential schedule slippage can
be detected and  possibly prevented before the  slippage becomes
unrecoverable.

494    COST AND ECONOMICS
   SHORT TERM ACTIVITIES
          • USE 50/50 RULE
s8 WEEKS
            50%
                                          — 100%
   LONG TERM ACTIVITIES          >8 WEEKS

          •  USE LINEAR BY DURATION RULE.
                                            100%
             0%
                        Figure 3
                             e MซAOOF
C BEHIND      I
  SCHEDULE    I
  UNO€R      |
  euOOET
 EAHNEO
 WILLIE
                      Figure 4
    Earned Value Data for a Typical Work Assignment

-------
  As stated earlier, the Earned Value variances are not absolute.
They are indicators of potential problems. They allow manage-
ment to evaluate trends on large work assignments  and on the
program as a whole. They are used to guide management in making
timely  and relevant decisions concerning the cost and schedule
elements  of the program.

WORK ASSIGNMENT EXAMPLE
  Figure 5 shows Earned Value Data Curves for  one REM III
Work Assignment. This assignment is  statused through  July 31,
1987. There are four periods over the  11 months shown for this
assignment where different conditions have affected performance.
Those  periods are indicated on Figure 5 as A to D.
 A    S    0    N   D   J


0  BUDGET
                               F r wen
                       +  ACTUALS
M   A   U   J   J   A   S


    o   CMWCD VALUE
                            Figure 5
                   Marathon Battery II  RI/FS
   Period A (September 1986 to December 1987) covers the initial
 activities on this work assignment. It is apparent from the curves
 that the work in this period was on schedule but consistently over
 budget. The situation was one of performing non-budgeted work
 in addition to the planned work.  This increased the actuals, but
 left the Earned Value unchanged. The actual effort was responding
 to the U.S. Bureau of Mines request to explain why certain innova-
 tive technologies were not recommended for Area I, a subject of
 a previous RI/FS. This unanticipated response to questions is the
 cause of the over'run shown. When planning a work assignment,
 it is important to consider carefully all possible work that may be
 involved. However, using Earned Value will assure that if unan-
 ticipated work is encountered it will  be immediately apparent to
 all concerned.
   During Period B (December 1987 to March 1987) the cost over-
 run worsened and  then improved slightly. In terms of schedule
 progress,  some  schedule  slippage  occurred in January and
 February. However, by March the project is shown to be slightly
 ahead of schedule. This period represents significant change to this
 work assignment. The cost over-run worsened due to responding
 to the state comments on the final work plan, health and safety
 training for the drillers required by SARA, a protracted interface
 with the drilling subcontractor due to project funding limitations
and a request to do a separate analysis for a separate RI/FS for
the plant area only. None of these activities had been budgeted
in the Work Plan. This situation was also the cause of the slight
schedule slippage noted on the curves.
  Period C is the time'frame when the site  investigation was
proceeding. During this period the work assignment was ahead of
schedule and under budget. The assignment was ahead of schedule
due to the U.S. EPA's request to perform limited field work on
one area of the site. In essence, this represented a change in the
technical baseline as compared with the original concept; in this
case, it is a down-scoping. The under-run is attributable to savings
in subcontracting costs, elimination of two wells from the scope
of the field  work and efficient performance of the field work since
no difficulties were encountered. All of these factors contributed
to the portrayal of a project that at the end of Period C is shown
to be under budget and ahead of schedule.
  The flattening of the budget curve in Period D shows that the
project was essentially on hold due to funding limitations. Some
work was continuing. This work involved an additional round of
sampling requested to verify the results of the initial sampling. Also,
questions concerning Area I, (the subject of a previous assignment)
were again being asked by the U.S. ACE.
  Our conclusion from this example is that at every point along
the  progress  of this  work  assignment,  the  Earned  Value
methodology accurately depicted the status with regard to the origi-
nal baseline. Additionally, it illustrates quite clearly the kinds of
problems and the various effects that those problems can have on
the execution of a work assignment.

CONCLUSIONS
  In conclusion, Earned Value is the only objective way to measure
progress. All other methodologies require the use of someone,s
subjective opinion in evaluating the fraction of completion of a
work assignment. The basis of Earned Value is the budget estab-
lished for all work on the program. This basis is quite logical and
provides the means to summarize progress  across work assignments
and regions and could well be used on the Superfund Program as
a whole.
   Earned Value focuses and sharpens management's attention to
cost  and schedule adherence. Not all cost and schedule problems
can be corrected. However, by focusing attention on these aspects
of the program, serious and chronic problems can be averted.
Furthermore, Earned Value encourages self discipline in the matter
of cost and schedule control at all levels  in the project organiza-
tion. As Earned Value is utilized on a program, cost and schedule
concerns become a part of the thinking process at the site manager
level—which is the level at which cost and schedule commitments
will be made or serious problems will develop.
                              Footnote
                                      This material has been funded wholly by the U.S.
                                    EPA under  Contract No.  68-01-7250 to Ebasco
                                    Services  Incorporated. It has been subject to  the
                                    Agency's review, and it has been approved for publi-
                                    cation.  Mention of  trade  names  or commercial
                                    products does not constitute endorsement or recom-
                                    mendation for use.
                                                                                              COST AND ECONOMICS    495

-------
                                      Bidding and Awarding  of
                            Remedial  Construction  Contracts  for
                                         Hazardous Waste  Sites

                                                  Patrick  F. O'Hara
                                                  Beth  F. Cockcroft
                                                   William C.  Smith
                                           Paul C. Rizzo Associates, Inc.
                                              Pittsburgh, Pennsylvania
ABSTRACT
  Superfund cleanup contracts are often competitively bid in order
to comply with established procurement policies and to create an
atmosphere of competition with the goal of completing the remedial
cleanups in the most cost effective manner practical.
  The organizations and facilities with the capabilities to under-
take a given remedial construction contract are often very few in
number. In particular, the scarcity  of facilities legally permitted
and technically qualified to dispose of waste from Superfund sites
makes a truly competitive environment  very difficult to achieve
for  many cleanup construction programs.
  The purpose of this paper is to assess  the procurement process
for competitively bid remedial construction contracts. Recommen-
dations regarding procurement policy are presented that will enable
the designers and agency personnel to obtain more responsive bids
by developing design packages  in  a manner that encourages a
greater number of responses and lower bids.

INTRODUCTION
  The procurement process for remedial construction projects con-
sists of the following principal steps:

  Preparation of a design and bid  package
  Establishment of bidder qualifications
  Advertisement  of the project
  Bid review
  Award of the construction  contract

  The following sections describe these steps and include recom-
mendations for performing each of these steps  such  that the
number, quality and responsiveness of the bids received are to the
contracting agencies' or organizations'  benefit.

PREPARATION OF THE DESIGN AND BID PACKAGE
  The format and organization of bid packages should be prepared
in a manner that facilitates their review by the prospective bidders.
One way to accomplish this is to utilize standard formats for the
preparation of engineering specifications and the technical aspects
of the bid package. Both the U.S. Navy (Navy) and the U.S. Army
Corps of Engineers (Corps) have approved a standard format used
for all construction projects for  which those agencies are respon-
sible.  This specification  format, with some adjustment, also is
applicable to hazardous waste remedial action contracts and has
been  utilized  successfully on remedial contracts  led by those
agencies for the  past  5  years.  Experienced bidders tend to be
familiar with these  types of  packages,  thus,  it  is highly
recommended that the formats  adopted by the Navy and by the

496     COST AND ECONOMICS
Corps  be utilized.  In  addition, guide specifications already
approved for many types of construction activities (such as fill
placement, sewage tine installation, water line installation and con-
crete construction) are published by these agencies and are availa-
ble  for public use. The utilization of these guide specifications
makes the bid package relatively consistent with past practices that
probably are familiar to prospective bidders.
  These guide specifications are edited by the design team in order
to make the remedial design site-specific and consistent with the
needs of the  given cleanup. The basic content and the  format of
these guide specifications have been utilized on numerous projects
within the hazardous waste, civil and military construction indus-
tries. Their utilization to the maximum degree practical is recom-
mended for  hazardous  waste  cleanup projects.  In addition, to
enable  the bidders to fully understand the construction intended
by the design documents and to perform accurate cost estimates
and quantity  "takeoffs," the  following should be considered:

• Use of commonly recognized construction symbols
• Normal construction  nomenclature
• Drawings of sufficiently large scale to understand the project
• Uniformity with respect to standard construction details

  These items, to which the construction industry (including the
remedial construction industry) has become accustomed,  will
provide an understandable design and will aid contractors in pre-
paring  their  bids.
  An excellent way to improve the clarity of the design and bid
package from the perspective of a bidder is to require the designer
to perform a contractor's estimate using the design and bid pack-
age. This cost estimate should be performed by a group of individu-
als  who did not participate in  the preparation of the design and
bid package. During the process of preparing this estimate, the
estimating team members should comment on the clarity and con-
structibility of the remedial design as presented  in the bid pack-
age. This type of quality control provides a good check of the design
and bid package prepared by the designer and should include pro-
visions for the resolution and disposition of the comments offered
by the estimating team as a normal part of the design review  process
to be implemented for hazardous waste remedial action projects.
  Another feature of the design package is the preparation of a
bid form that distinctly describes the overall scope of work in a
manner that is logical from a  bidding standpoint. The bid form
should include both lump-sum and unit price aspects of the project,
as appropriate and in all cases should be described in detail in a
measurement and payment section of the specifications. The
measurement and payment section should describe in detail the

-------
physical aspects of the remedial action program in exactly the same
manner as presented on the bid form. The intent of a measure-
ment and payment section of the specifications is to clarify which
aspects of the design are to be included in each line item of the
bid form. In addition, it is highly recommended that every bid form
include a line item at the bottom for direct entry of a price by the
bidder to encompass "all other work required by the contract docu-
ments not included in other bid items." The purpose of this specific
line item is to compel the prospective bidders to review in detail
the design and bid package and to reduce the possibility of subse-
quent claims during the remedial construction process. This require-
ment helps eliminate claims  which originate as  a  result  of
inconsistencies between the requirements of the bid form and the
technical requirements of the plans and specifications.
  Another important feature for remedial design and bid pack-
ages is the explicit provision for variation in  quantity clauses.
Because  of the uncertain nature of the quantities associated with
many remediation projects, up-front mechanisms should be estab-
lished in  the contract during the procurement phase to handle large
variations and quantities. A fair variation in quantities clause will
reduce bidder's contingencies for the unit price bid items most likely
to  be affected by the large variation and estimated  quantity.
Examples of such line items are excavation and removal  of con-
taminated soil on a volume or  tonnage basis and treatment and
discharge of contaminated water on a volumetric basis.
  An additional feature of the  design  and  bid  package  for
hazardous waste remedial action projects should be  the inclusion
of a concise but very complete description of all chemical analysis
data available to the design team regarding the site. This informa-
tion is important to the prospective bidders, both from the stand-
point of technical understanding of the site as well as the bidder's
need to consider such data in the development of site-specific health
and safety programs. Not including such information in a design
package imposes  severe risk with respect to liability on both the
designer and the  owner. In addition, it is highly appropriate to
include all available geotechnical information such as boring logs,
cross-sections, physical laboratory data, well installation diagrams
and water levels as an appendix to the design package. This infor-
mation is included to reduce contractor uncertainty and to mitigate
future claims against the designer and/or the owner with respect
to withholding site information that could have been important
to the contractor in the performance of the remedial activities at
the site.  It is recommended that these data always  be  included.
  A special requirement of Superfund projects is the clear deline-
ation of the technical specifications of the explicit off-site disposal
requirements  for wastes  considered  hazardous that are being
removed off-site for ultimate disposal. Standard contract language
that has been utilized and approved by the various regulatory
agencies is available that describes in legal terminology the U.S.
EPA's position on off-site disposal policies for Superfund sites.
Clear definition of this policy will help preclude future claims result-
ing from a vague definition of this policy in the design and bid
package.
  A final and probably most important point of discussion in the
preparation of a design and bid package for remedial cleanup is
the development and specification for remedial activities that will
permit new, alternative and/or competitive technologies (sometimes
proprietary technologies) to be  utilized in the cleanup effort. The
means by which alternate technologies can be considered in a com-
petitive bid package are:


• Utilization of performance  specifications in lieu of method
  specifications for on-site waste treatment or off-site waste
  treatment
• Permitting alternate disposal methodologies from those speci-
  fied in the design cleanup package
• Specifying alternate remedial design features and requiring bids
  on each of those specified features
  All three techniques have been successfully employed on public
competitively bid remedial design projects under the Superfund
Program. Examples with which the authors have had personal and
specific experience are the preparation of bid packages for both
synthetic and natural clay caps for waste disposal sites. These pack-
ages require the contractor to quote on one or both of the capping
alternatives with the bid package including complete design infor-
mation for each alternative to be quoted. Another example was
providing the option for on-site treatment and recharge  of exca-
vation water or off-site transportation and disposal of that same
water. The design package gave performance specifications to be
met for the on-site treatment and recharge option and cited the
standard off-site disposal requirements for off-site transportation
and disposal. A third example was the use of performance specifi-
cations for a permanent on-site leachate treatment facility during
the post-closure period at a Superfund site.  The specifications
provided quantities as  well as ranges of contaminant concentra-
tions for the influent leachate to be treated and cited both average
and maximum discharge limits for the treated effluent. The con-
tractor will be required to demonstrate acceptable performance with
respect to the performance specifications during a trial operational
period and  is responsible for selecting and demonstrating  the ade-
quacy of the technologies that the contractor has chosen to employ
to treat the specific waste stream.
  These types of technical criteria delineations in the design and
bid package enable increased competition and the introduction of
alternate or proprietary technologies to the competitive bid  process.

ESTABLISHING  BIDDER QUALIFICATIONS
  It is recognized  both in the public and the  private sector that
the cleanup of Superfund sites  should  be accomplished  by con-
tractors who have the expertise, the experience and the resources
to properly do the job. Under long-term remedial type construc-
tion projects, contracting organizations historically have developed
several mechanisms to  assure or at least maximize the likelihood
that a qualified organization will be contracted  to perform the
cleanup. Many  competitive  procurements   are  initiated by
developing a short list of qualified bidders based upon bidders
establishing credentials of historical performance and  demon-
strating that performance by fulfilling pre-established criteria that
are indicative of their ability to undertake the work. Oftentimes
bidders must complete a prequalification document prior  to being
awarded a  bid package that will enable them to compete for the
project on  a cost basis. In effect, this technique results in a two-
phase competitive procurement. The first phase is  a phase of com-
petitive procurement independent of cost considerations for the
specific contract being considered. The second phase includes only
those bidders deemed to be sufficiently qualified to perform the
work  and puts those bidders into a price-type competitive bidding
situation in which award will be based  solely on the least cost to
the owner,  as perceived by the owner. The advantage of this two-
step procurement process is that the second phase of procurement
can be  a "black and white" decision  in which  the contracting
agency is simply able to award the project to the lowest bidder that
has already demonstrated credible qualifications to undertake the
project. This type of procurement practice is  typical of that em-
ployed by many organizations to procure professional services, in-
cluding design services.
  A second type of pre-selection results from establishing signifi-
cant bonding requirements for  the performance of the remedial
contract. Bonding organizations have  become increasingly cog-
nizant of the special nature of hazardous waste remedial construc-
tion  and often are excellent  at  sorting out and  eliminating
unqualified contractors from the procurement process. Some recent
government procurements for hazardous waste remedial construc-
tion projects have utilized significant bonding as  the sole delinea-
tion of the qualified bidders.
  Agencies in the private sector should utilize a combination of
establishing and verifying technical and  financial credentials of
                                                                                              COST AND ECONOMICS    497

-------
bidders prior to awarding construction contracts and should utilize
prudent bonding practices to assure that those entities with which
they contract are capable of fulfilling the goals of the remedial con-
struction program. In all cases, it is recommended that a clear series
of mechanisms be established to determine qualified versus non-
qualified bidders. The final phase of the selection process prefer-
ably should be based on simple selection of the most financially
advantageous bid submitted by those organizations already demon-
strated to be qualified to execute the construction contract.

ADVERTISEMENT  OF THE PROJECT
   In this context, advertisement is defined as the actual letting of
the bid package to those organizations who are potentially respon-
sible bidders for the  contract. The time period  allocated to the
bidders for review of the bid package and the preparation of their
bids is of considerable significance. Allowing insufficient time to
prepare hard-money quotations for performing a hazardous waste
construction project will normally lead either to excessive contin-
gency  in the bidders' price quote or inadvertent omissions from
the bidders' own construction cost estimate. These omissions may
induce the bidders to  either reduce cost with a possible reduction
in the quality of the services provided or to submit additional claims
to cover the omissions made in the original bid  or cost estimate.
   The allocation of too  much time for the review and the submit-
tal of a bid may result in  over-refinement of the bid by the remedial
contractor with a corresponding increase in the conservativeness
of the  price quoted. This second aspect is somewhat mitigated by
the frequent practice in the real world of waiting until the last week
that bids are due to prepare the cost estimate for the remedial con-
struction project. However, a judicious selection of the time period
allocated for review and submittal of bids is very appropriate. Good
practice dictates that this period should range from 30 to 45 days,
depending upon the complexity and the likelihood of extensive sub-
contracting of the given  remedial construction program. If exces-
sive feedback and  numerous requests for clarification on the bid
package are received  by the contracting organization during the
bid review process, the issuing of amendments to the bid package
as well as the extension of the bid date are always appropriate.
The  allocation of insufficient time for this entire process  only
decreases  the number of responsive bidders and  increases the
remedial construction costs ultimately paid by  the contracting
organization.

BID REVIEW AND  AWARD
  The  review and award process will  be greatly simplified  if a
systematic bid form has been utilized and if bids have only been
 received by contractors who were prequalified either by bonding
 or by preliminary technical prequalification. If bidders modified
 the bid form or submitted alternate proposals and methodologies
 not explicitly called for in the design documents,  it is strongly
 recommended that these alterations be carefully and constructively
 reviewed and not simply be used as a means of eliminating an other-
 wise responsible and responsive bid. Normal contract language in
 construction procurement packages will give the reviewing agency
 the option  to discount such non-standard bids. However, it is
 always in the contracting organizations' best interest to seriously
 review alternate and modified proposals. It is recognized that pri-
 vate entities have considerably more flexibility in this regard than
 government-sponsored procurements. The private agency should
 take advantage of this difference and view alternate proposals con-
 structively.
  Upon review of  all proposals and the pricing submitted, the
 apparent qualified low qualified bidder should be denoted and noti-
 fied. A meeting should be held with the apparent low bidder in
 order to discuss any misunderstandings that the bidder may have
 with respect to the technical, legal or financial requirements of the
 anticipated contract. It may be decided at this meeting that it would
 be in the best of interest of both parties if the apparent low bidder
 did not proceed with the project.
  If a highly-qualified apparent low bidder has been identified,
 it is in the best interest of the contracting organization to solicit
 value engineering type proposals  on the design as a means of
 determining if the bidder has an alternate scheme,  methodology
 or technology that was not reflected in the original proposal. Actual
 execution of the  remedial construction contract would normally
 take place if both parties mutually agree that they want to enter
 into the contract at the conclusion of this meeting.


CONCLUSION
  The  above sections of this paper cited experiences and  recom-
mendations of the authors for improved efficiency and satisfactory
execution of the process of acquiring a remedial construction con-
tractor for the cleanup of a Superfund site. These recommendations
are an  outgrowth of both normal good practice in the construc-
tion industry and some of the special and unique characteristics
of the hazardous waste remedial cleanup industry. Means of con-
tinuing  to improve the solicitation and performance of remedial
construction contracts will continue to evolve. The authors grate-
fully acknowledge the clients, the regulator, agencies, the contrac-
tors and the colleagues who have added to our experience in the
remedial cleanup  process.
498    COST AND ECONOMICS

-------
                             Environmental  Risk  Considerations
                                       In  Real Estate  Transfers
                         For Active  Waste  Management  Facilities

                                                 Mark  P. Zatezalo
                                                 Patrick F.  O'Hara
                                          Paul C.  Rizzo Associates, Inc.
                                              Pittsburgh, Pennsylvania
ABSTRACT
Technical evaluations, environmental audits and environmental risk
assessments are routinely performed as part of the overall process
of property ownership transfer when the site history includes any
past or suspected usage of hazardous materials.
  Of special interest is the assessment of environmental risk for
acquisition or divestiture of active waste management facilities.
These sites are a very specialized class of industrial facility, and
a process of evaluating environmental risks on behalf of buyers
and indirectly sellers is described.  The methodology described
herein represents an attempt to rationally approach the purchase
of a facility which has some inherent risk which may or may not
be manageable.

INTRODUCTION
  Conducting business implies accepting and dealing with risk. If
there were no risk attached to business transactions, there actually
would be no reason for business and no such terms as "success"
or "failure" in business development. Today, part of the risk one
assumes in business is the risk associated with purchase of another's
property. Assessment of environmental risk is a new variable in
an ever expanding list of variables that one must consider when
purchasing new  facilities. And herein is a paradox consisting  of
the following elements:

• The business entity may wish to obtain the necessary property/
  facilities to expand its revenue base and profits
• The business entity does not want to assume a liability due to
  property acquisition that will make the profits it hopes to achieve
  not worth the investment
• The purchasing entity  may not  have the time and  money
  resources required to study the site to the degree that the poten-
  tial environmental problems are strictly quantifiable

  This paper is an attempt to deal with property transfer issues
from the buyer's perspective. The procedures outlined are intended
to follow a declaration or finding that some risk does exist and
actually forms an assessment of what the cost/benefit ratio is likely
to be if the property is purchased in spite of the risk. An area of
the economy where the assessment of risk is of particular impor-
tance occurs in the transfer of waste management facilities.
  When dealing with potential acquisition of waste management
facilities, an appropriate assessment of environmental risk assists
in minimizing the amount of guess work as to environmental
exposure when acquiring property and aids in the decision-making
process. The degree of risk assumed by the buyer is a function of
both the uncertainty associated with the property itself and the ad-
ditional uncertainty inherent in the assessment process. Both areas
of uncertainties must be characterized and understood.
  The proper assessment of waste disposal facilities prior to
purchase is a task for private business hoping to expand as well
as municipal and state authorities wishing to establish more control
over  local  waste  management activities.  For various reasons,
purchase of existing facilities can be very attractive if liability issues
can be better understood and a comfort level can be obtained.
  The optimum decision-making team for assessing  potential
acquisition sites is shown on Figure 1. The structure of the body
is rather simple, but the importance of involvement of each element
must be stressed.
                         Figure 1
           Decision-Making Team For Risic Assessment
   Obviously, upper level management takes the lead in final deci-
 sions. They must weigh all issues and decide whether the risk levels
 involved in the acquisition  are  acceptable.  To make sound
 decisions, management must draw on the experience and judgment
, of both technical expertise (e.g., engineers, hydrogeologists, etc.)
 and cost analysts to provide important input to the decision-making
 process.

 FACILITY SITE CHARACTERIZATION
   When dealing with waste management facilities, the principal
 uncertainty concerning the property transfer is associated with the
 property from the "ground down." The techniques of assessment
.discussed here focus on the subsurface migration area of facility
                                                                                        COST AND ECONOMICS     499

-------
risk. Conventional methodologies that are appropriate to charac-
terize risks from the "ground up" should be employed to assess
that aspect of facility risk in conjunction with the "ground down"
assessment described below.
  The first step in any investigation dealing with site characteri-
zation is to review all site data that can be obtained within the
allotted time  period.  Important data that are typically  available
can be used to formulate an understanding of site and regional
geology:

  Site topography and physiography
  Geology and hydrogeology
  Surface water hydrology
  Surface water and ground water quality
  Climatology
  Facility engineering plans

  Typical buyers generally want as much information as possible
under severe  time schedule constraints; unfortunately, resources
to be expended are usually at a premium. Thus, it is very impor-
tant that experienced individuals conduct the investigation and that
the individuals limit the scope of their conclusions and properly
qualify any opinions rendered commensurate with the scope of the
assessment.
  The data obtained in the areas noted in the foregoing list can
be used to develop an understanding of the physical setting or model
of the site. This site model should note points of water inflow and
outflow to the property; it will be used to estimate flux rates of
water through the site and the potential  for surface and subsur-
face migration of contaminants from the site.

Characterize  Nature of Contaminants
  The  nature and extent of  the substances of concern must be
assessed. Time and resources  may or may not permit an analytical
confirmation of the background information, but in all cases, it
is imperative to characterize the substances of concern. This charac-
terization has direct implications when dealing with transport and
fate in the environment and potential environmental risk. In most
cases, the exact constituents will be difficult to determine, but some
key parameters will aid in the analysis. These parameters include
but may not  be  limited to the following:

  Specific conductance
  Chlorides
  Total organic  carbon
  Volatile organics
  Biological oxygen demand
  Chemical oxygen demand

  If available, these parameters  can give  an indication  of the
amount of water quality degradation (if any) that has taken place
in the vicinity of the site.

Determine Pathways of Migration
  Having constructed  a general hydrogeologic  framework or
model; the potential migration pathways are determined. Using
a non-numerical model of the facility site, all potential pathways
should be identified. Again the investigator's experience must some-
times replace field work and subsurface  investigations  as the
primary analytical tool, when time and resource commitments are
short. At this point, a  person who can easily  "visualize" the
microtransport mechanisms in the  subsurface is most valuable.
Potential pathways of migration should be  traced to areas where
impact to health and/or the environment are likely.

Determine Points of Risk
  Using pathway studies and assessment of migration potential,
points of risk must then be assessed to determine areas where the
threats to human health and the environment are possible. The
greatest such points of risk may include, but are not limited to:

• Downgradient wells
• Downgradient streams
• Surrounding air
• Points of direct (physical) contact

  Examples of receptors in these areas include people, wildlife,
fish and livestock. It is important to identify the most immediate
points of risk where the most acute exposure will occur. These areas
are used to form  the basis for determining remedial alternatives
later in the assessment  process.

"WORST CASE" AND "MOST PLAUSIBLE
CASE" SCENARIOS
  Upon developing an  understanding of site conditions through
site characterization, the next step is to develop "worst case" and
"most plausible case" impact and remediation  scenarios con-
sidering the nature of the substances of concern, the migration path-
ways and the nature and  location of the points of risk. These
scenarios have a corresponding level of remedial effort that may
be necessary in  the  future.
  The "worst case" assumes that remediation  required must
eliminate contaminant migration and renovate groundwater in areas
already  contaminated.  The "most plausible" case  takes into
account the location of downgradient receptors and the perceived
need to mitigate existing migration. Factors used in this analysis
include regulatory philosophy, downgradient risk, population den-
sity and visibility of the facility being evaluated. The rationale for
the level of mitigative effort required for a "most  plausible" case
can be assessed.
  It  should be  stated that,  given the  shifting  of regulatory
philosophies and the ever increasing public awareness of ground-
water pollution, the assessment weighted toward the "worst case"
scenario is probably prudent.
  The exercise of testing the difference between "worst case" and
"most plausible" case  is meant to be a sensitivity test to deter-
mine the difference in remedial cost between "most plausible" and
"worst" case scenarios and forms an  estimate of the range of
remedial cost one may incur.

ASSESS POTENTIAL REMEDIAL ACTIVITIES
  At  this point,  the decision-making mechanism is  placed in
motion. Upon completing the scenarios outlined above, some ideas
must be formed as to the types of action required to remediate
the problem and  the cost associated with the potential remedia-
tion. The assessment of potential remedial action obviously is the
key to  determining potential detrimental  environmental and
economic impacts due to existing site conditions.
  Using background data, it is up to the engineer/scientist/esti-
mator to determine the cost of potential remediation. If ground-
water degradation is encountered at a point of environmental risk,
such items as the cost of the containment, collection and treatment
of contaminated water, physical removal of the contaminant source
and other remedial efforts must be estimated.
  Accurate estimation of costs is critical to the assessment of im-
pacts to the decision to acquire or not to acquire the property. Also,
the method of payment and amortization schedule have large im-
plications as to feasibility of purchase.
  Possibly the most delicate step in the operation arrives when,
after the most likely remedial action and associated cost is deter-
mined, a probability that the remediation will be necessary must
be calculated. This step, nebulous at best, involves all sections of
the decision-making team and requires a knowledge of the site area,
regulatory philosophy at the local, state and federal level and the
perceived threat to  human health and  the environment.

A CASE  HISTORY
  The example chosen (Figure 2) is an operating landfill in this
country. Approximately 1 mile away and downgradient is a lake
used for recreation. Several houses are located crossgradient (north
and south) from the landfill. Also located adjacent to and some-
500     COST AND ECONOMICS

-------
what crossgradient from the site is a former chemical disposal site
now on the NPL. The site stratigraphy consists of 10 to 30 ft of
glacial deposits underlain by sedimentary bedrock (Figure 3). Also,
immediately downgradient from the landfill is a closed municipal
waste landfill.
           GROUND WATER
             . FLOW
                                          r LANDFILL SLATED FOR
                                           POSSIBLE ACQUISITION
                                                    D  HOUSE
                                        Q HOUSE
                            Figure 2
           Site Plan of Proposed Property to be Purchased
           '?'   *     ฐ GLACIAL  DEPOSITS  .o -4.
  *.
  •o.
'ซ•    *.
                                                    B/-.0
                                                     ~ '  A .
                                                       .i •
                           Figure 3
     Typical Cross-Section of Proposed Property to be Purchased
 Hydrogeologic Characteristics
   Using site-specific data pertaining to the landfill and data col-
 lected while  studying the adjacent NPL  site, several  pertinent
 preliminary findings were obtained including:

 •  There is at least one saturated zone in glacial deposits recharged
   by precipitation and  surface water impounded in upgradient
   areas by earthwork associated with the landfill
 •  Saturated zones also exist in bedrock underlying the site down-
   gradient
 •  Groundwater levels in the bedrock are less than 2 ft lower than
   those in the glacial deposits
• Water quality from these zones show minor amounts of impact
  from site and/or nearby operations
• Water quality at the NPL site shows that the shallow ground-
  water exhibits medium levels of contamination while deeper
  zones show minimal impact

  This framework is now used to establishing pathways, points
of risk and finally remedial  scenarios.

Determination of Pathways  to  the Environment
  Based upon data collected, the main pathways to the environ-
ment consist primarily of flow through the shallow saturated zones
to downgradient surface water bodies. To a lesser extent, the deep
bedrock flow zone  is also a  probable flowpath.

Determination of Downgradient Risk
  Expected present risk in this case is currently low due to the rapid
attenuation of  contaminants observed in  the  monitoring wells
adjacent to the  landfill and the lack of downgradient risk to the
west of particular importance. Thus, the main downgradient point
of risk, the lake, is expected to be at relatively low risk at this time.

"Worst Case"  and "Most Plausible Case" Scenarios
  The "worst case" scenario envisioned was a total containment
scheme with renovation of groundwater in the site vicinity.  Ele-
ments  in the  remedial scheme include:
  Grading and  placement of cap
  Surface drainage control
  Active retrofit leachate collection and treatment
  Placement  of slurry wall
  Placement  and pumping of groundwater  collection wells  for
  renovation purposes
  Preliminary costs for  performing  these remedial tasks were
developed and the most plausible case was then assessed.  Since there
were other contributing factors to environmental degradation in
the site vicinity and present impacts of the landfill to groundwater
appear minimal, it was assumed that groundwater renovation work
would not be necessary or would be attributable to other sources
and the minimization of leachate generation and containment on-
site was the main required mitigative measure. The most plausible
remedial efforts therefore are probably:
• Grading and  placement of the cap
• Surface drainage control
• Placement  of slurry wall
• Retrofit leachate  collection system  (designed to develop an in-
  ward gradient to the landfill, thus minimizing migration of con-
  taminants)

  Based upon costs associated with these remedial efforts and the
anticipated landfill revenues  over an expected life of 20 years, it
was decided that the risk  and associated cost of remediation was
within the acceptable limits and the facility was purchased.

CONCLUSION
  Many serious organizations too often choose to walk away from
a potential property acquisition because of environmental risk. In
cases where the  risk may be acceptable, an analysis should be per-
formed to determine the economic viability of  purchasing  a
property in spite of the fact that some potentially significant risk
may be apparent. The description of methodology outlined in this
paper to study waste management facilities that may be potential
acquisition candidates should be seen as a first step  in  the process
of assessing property transfer risk from the purchaser's point of
view.
  The scope of the assessment should be a function of the initially
perceived environmental risks as well as the initially perceived eco-
nomic benefits should acquisition be completed. Using the metho-
dologies described, a rationale can be developed that will allow
for proper planning of business expansion while taking into account
the potential risk factors.
                                                                                              COST AND ECONOMICS     501

-------
                     A  Demonstrable  Performance Approach  for
                        Waste Isolation  in Geologic  Repositories

                                            Daniel  C. Melchior, Ph.D.
                                                  Sarah Hokanson
                                       The Earth Technology Corporation
                                              Long Beach, California
                                                Alexandria, Virginia
                                                   Jon Greenburg
                                     U.S. Environmental Protection Agency
                                               Office of Solid Waste
                                                 Washington,  D.C.
ABSTRACT
  The disposal and isolation of untreated CERCLA and RCRA
waste in geologic repositories is currently being considered an eco-
nomically and environmentally attractive alternative to waste treat-
ment and disposal  in surface  land disposal units.  To  assure
long-term isolation,  site-specific data and numerical simulations
of the geohydrology, geochemistry and geomechanics provide the
best method to estimate and predict repository performance.
  A conceptual model of in situ geohydrology, geochemistry and
geomechanics of the natural formation must be initially formu-
lated and numerically analyzed. Then, once engineering design and
repository sealing approaches are established, post-closure per-
formance should be evaluated. Analyses should predict how the
disturbed  zone, engineered systems and repository seals will behave
and whether predetermined performance criteria of no release can
be met.
  A two-fold parallel strategy is proposed to meet these goals. One
effort evaluates the detailed performance of repository components
while the other evaluates total system performance using coupled
simplified numerical functions derived from the more complex
component models. Existing computer codes are proposed to evalu-
ate both component and system response.

INTRODUCTION
  Disposal and isolation of untreated CERCLA and RCRA waste
in geological repositories constructed in existing mined space is
considered an economically and environmentally attractive alter-
native to  waste treatment and disposal in surface land disposal
units. Geologic repositories conceptually offer long-term isolation
that effectively removes waste from the biosphere for  as long as
it remains hazardous. Geologic repositories of both hazardous and
moderately radioactive waste currently are being developed in salt
anticlines in Germany; in North America, over one thousand geo-
logic repositories are used to store information files and natural
resources. The interest in geologic repositories for radioactive waste
has heightened since the National Academy of Sciences recom-
mended them for disposing such waste in the U.S.1 The Nuclear
Waste Policy Act of 1982 further promulgated this concept into
law.
  Existing mines in  salt and limestone formations are thought to
be excellent candidates based on a variety of factors and are being
privately  pursued. The factors include depth, relative degree of
hydrologic isolation, size of underground workings, geotechnical
characteristics, regional geologic stability and deposit economic
value and scarcity. Mines for metallic minerals also are potential
repository hosts but typically do not have the necessary favorable

502    DISPOSAL/STORAGE
factors to make them as attractive. Figure 1 shows a map of the
location of mining districts that appear to be potential candidates
for hazardous waste isolation2. The close proximity of most
districts to industrial centers is another factor that would contribute
to a potential repository's economic feasibility.
                          Figure 1
        Map Showing Location of Mining Districts Surveyed
  Interest in using geologic repositories has led the Hazardous and
Solid Waste Amendments of 1984 to include provisions authorizing
the U.S. EPA to define performance and permit standards. New
regulations under Subpart X of 40 CFR 264 and 40 CFR 270
proposed on Nov. 7,1986 establish the permitting framework for
repositories to operate. Under current U.S. EPA regulations,
geologic repositories for untreated wastes must demonstrate that
no hazardous constituent migration to the accessible environment
will occur  for as long as the wastes remain hazardous.
  For a "no-migration" variance to be granted, a site-specific per-
formance evaluation of waste isolation must be conducted. To
accomplish this, a program that uses Held and in situ measurements,
laboratory experiments and numerical modeling to predict how
natural or man-induced processes will influence repository perfor-
mance is needed. The methodology presented here  provides a
demonstrable  performance approach for waste isolation in geo-

-------
logic repositories post-closure using all these data sources provided
from geologic, hydrologic, mechanical and geochemical studies of
the site and generic sources.
  This post-closure criterion is constrained by the "no migration"
issue. Since a proposed repository cannot demonstrate before con-
struction that it can qualify for a "no-migration" variance, a
methodology  is needed to assess system release. System release
depends on four performance objectives whose interrelationship
(Figure 2) will govern qualification for a no-migration variance.
These performance objectives include site geology, groundwater
flow, engineered containment structures and repository seals. The
methodology  proposed here to  demonstrate post-closure reposi-
tory performance relies on assigning expected performance objec-
tive criteria for global processes which in turn must be evaluated
to identify performance goals and error ranges for the various com-
ponents, such as the waste containers or seals.
     ENGINEERED
     CONTAINMENT
     STRUCTURES
                             Figure 2
             Performance Objective Interrelationship to a
              Geologic Repository System Release Study
                                               Once these goals and values are established and deemed achiev-
                                             able, a global post-closure performance assessment is performed
                                             to identify if the no-migration regulatory constraint can be met.
                                             This methodology  will  provide a  demonstrable performance
                                             approach  for a permit to isolate waste in geologic repository.

                                             POTENTIAL GEOLOGIC REPOSITORIES IN
                                             EXISTING NON-METALLIC  MINES
                                               Geologic repositories in underground mines must demonstrate
                                             long-term structural stability and waste isolation capability. These
                                             performance factors depend on geologic  and hydrologic charac-
                                             teristics of the host rock formation and design and operating factors
                                             of the mined space. Identifying candidate sites for geologic reposi-
                                             tories therefore involves an initial screening of suitable host for-
                                             mations, followed by an evaluation of the suitability of the mined
                                             spaces within the selected formations. The criteria for suitable for-
                                             mations and mines are discussed below,  along with site-specific
                                             data requirements for performance demonstration.

                                             SUITABLE FORMATIONS AND MINES
                                               The  most suitable candidate  formations and mines meet the
                                             following siting criteria. In general, formations that have not been
                                             previously mined are not considered economically viable unless they
                                             can be readily developed (i.e.,  solution mining in salt domes).
                                             Similarly, formations located below 1,500 ft are uneconomic and
                                             structurally unattractive. A minimum host formation thickness of
                                             100 ft is essential to ensure structural integrity of the mined opening.
                                             Surface and subsurface  coal mines  are not considered suitable
                                             mined spaces because of the fire hazards, high-fracture density and
                                             unworkable mine openings. Caves  generally are poorly charac-
                                             terized in  terms of  geology  and hydrology and probably would
                                             require extensive underground construction. Finally, those forma-
                                             tions located within 100 miles of "Standard Metropolitan Statistical
                                             Areas" (SMSAs) in the contiguous 48 states are considered more
                                             attractive.
                                               Formations within 16 mining districts could meet the prescreening
                                             criteria listed above. In addition, the following geologic and hydro-
                                             logic properties would influence repository performance:

                                             • Depth
                                             • Hydrology
                                             • Regional geologic stability
                                             • Geotechnical and rock mass characteristics
                                                               Table 1
                                                Potential Geologic  vs. Suitability Criteria—
                                                      Non-Metallic Mining Districts
                                    (The generalized location of these deposits may be found in Figure 1.)
Criteria
Mine Type
Range 1n Depth of
Excavated Area
Bedded Salt and Potash
Shaft/Room and Pillar
Most mining 1s at depths
less than 1,500 feet,
commonly 1,000 feet
Salt Domes
Shaft/Room and Pillar
Usually between 200 and
2,000 feet
Limestone
Room and Pillar (Cathedral-like)
Varied depths, usually less than
1,000 feet
Illinois Fluorspar
Shaft/Room and Pillar
Up to 1,500 feet, typically
900 feet
    Hydrologic
    Considerations
Usually no flooding problems
except In over-extracted
mines
Usually no flooding problems
except In over-extracted
mines
Some areas subject to karstlfl-
catlon.  Some flooding 1n
certain urines
Some flooding 1n very deep
mines
    Geotechnical and Rock
    Mass Characteristics
Bedded salt and potash,
thicknesses of a few feet to
several hundred feet,  few
fractures  or discontinuities
other than bedding
Dlapiric salt—good rock
mass characteristics—few
fractures
Bedded limestones and dolomite.
Mined deposits usually massive,
bedded without extensive
fracturing
Vein filling In carbonate
rocks, fractures common
    Regional Stability
Michigan Basin -  stable
Appalachian - only minor
seismic activity
Permian   stable
Tectonlcally stable
D1ap1r1c movement could be
a consideration
Limestone  mining concentrated 1n
Midwest and East-Central U.S.
stable areas
No significant historical
seismic activity or
tectonlsm Identified
                                                                                                    DISPOSAL/STORAGE    503

-------
  In  general,  non-metallic mining districts  that  produce salt,
anhydrite, limestone and potash are better suited for hazardous
waste disposal than metallic mining districts, based on hydrologic
and geotechnical characteristics2.  Table 1 describes the general
suitability of these districts as a function of the specific criteria
identified2-3.
  Room-and-pillar or solution mining are most suitable for geo-
logic repositories because they: (1) rely on formation rock natural
support with little artificial support and (2) create minimal disturbed
zones. Suitable mine designs are those  that are structurally sound
and have large enough openings to accommodate access, waste
emplacement and  waste retrieval.

SITE-SPECIFIC GEOLOGIC AND HYDROLOGIC FACTORS
  The stability of a proposed  repository and the degree of waste
isolation are directly affected by the site geology and hydrology.
Regional and local geologic setting must be fully characterized to
assess repository performance. Principal geologic information to
be  considered in the  site characterization includes:

• Stratigraphy and mineralogy
• Structural and tectonic setting
•  Rock-mass  characteristics
• Other  subsurface exploration activities

  Stratigraphic information provides the geometric foundation for
three-dimensional conceptual models  from  which  quantitative
predictions of repository performance can be derived and provides
a general understanding of the regional mineralogy, structural and
tectonic setting and rock-mass characteristics.  Table 2 lists the
above  factors along  with specific information to be obtained,
characterization methods and limitations to the characterization4
                                                    While the above geologic features may be used to identify poten-
                                                  tial pathways of release, a hydrologic assessment is needed to assess
                                                  whether fluids can enter a repository via these release points and
                                                  transport wastes. Water  sources,  including groundwater from
                                                  aquifers, brine inclusions and surface water, should be identified
                                                  and located and their hydraulic and chemical characteristics should
                                                  be fully described. Table 3 identifies specific information needs,
                                                  characterization methods and limitations to the characterization.

                                                  DESIGN, OPERATION  AND CLOSURE OF MINES
                                                    Geologic repositories may involve a broad range of design and
                                                  operating practices depending on the type of mine. These design
                                                  and operating practices are critical to the issues of post-closure per-
                                                  formance. For example, solution-mined caverns in salt domes may
                                                  be designed to have a minimal  disturbed  zone in the host while
                                                  room-and-pillar mines typically have larger disturbed zones created
                                                  as a result of mining which might create enhanced fluid flow into
                                                  the repository.
                                                    Regardless of the type of mine, design and operating informa-
                                                  tion must be obtained under the following four general areas of
                                                  concern to assess post-closure repository  performance:

                                                  • Structural stability of design
                                                  • Waste emplacement  operations
                                                  • Waste characteristics and compatibility
                                                  • Mine closure seals


                                                    Table 4 lists these four factors and the relevant information to
                                                  be obtained. All of this information will be  needed to assess reposi-
                                                  tory performance response.
                                                                  Table 2
                                           Geologic Information for Repository Performance Analysis
          Property
                        Information Heeds
                                                  Characterization Methods
                                                                          Limitations to the
                                                                          Characterization
                                                                                                      Presentation of Data
          Stratigraphy
-Lateral and vertical
 extent of unit.
-Nature of the contacts
 Between units.
-Other Stratigraphic
 discontinuities.
-Rev1eซ existing geologic
 naps/cross sections.
-Field upplng.
-Surface geophysical
 survey.
-Structural complexity
 of  the region.
-Little structural/
 chemical centrist
 between units.
-Fen geophysical or
 borehole stapling
 points.
-Regional and site geological
 maps.
-Cross-sectional maps.
-Isopach taps.
          Structural     -Structural  dlscontl-
          and            nultles.
          Tectonic      -Regional  plate tectonic
          Setting        setting and activity.
                       -Present temperature,
                        confirming  pressure
                        conditions, pore fluid
                        pressures.
                       -In situ state of stress.
                       -Future 1n situ conditions.
                          -Field napping.
                          -Drill core and sampling.
                          -Geophysical surveys.
                          -In situ stress measure-
                           ments.
                          -Geophysical surveys.
                          -Modeling.
                        -Small-scale variability
                         of subsurface  structures.
                        -Few geophysical or bore-
                         hole sampling  points.
                        -In situ stress calcula-
                         tions Involve  large
                         errors.
                        -Sampling point
                         availability.
                        -Crrors In modeling
                         future conditions.
                         -Geologic amps.
                         -Structural fence diagrams.
                         -Rose diagrams or sterdonets for
                          surface and drill-core  fracture
                          orientations.
                         -Core Indices and fracture
                          histograms for each drill hole.
                         -Text discussion relating
                          tectonic setting to structure
                          and stratigraphy.
          Mineralogy
-Mineral content and
 chemistry.
-Petrologtc  and
 petrograpMc data.
                                                 -Laboratory studies.
                                                 -Geophysical  surveys.
                        -Solution-mined
                         o Few borehole  sampling
                           points near repository
                           site.
                        -Existing mine
                         o Poor definition of the
                           formation anay from the
                           mine Itself.
                         -Graphs and figures for
                          geophysical and  rock coring
                          data.
          Rock Mass
          Character-
          istics
                       -Stress-strain parameters.
                          -Laboratory tests.
                          -In situ tests.
                        -Inability to accurately
                         recreate In situ cond-
                         dlttons.
                        -Poor understanding of
                         1n situ stress mechanisms.
                        -Laboratory uncertainties.
                          -Stress-strain diagrams.
                          -Creep curves plotted.
                          -Creep measurements plotted
                          versus time.
          Other
          Subsurface
          Activities
          and Resources
-Number, location, depth,
 and present state of
 artificial  penetrations.
-Geologlc/nydrologlc
 resource data.
-Literature  searches.
-Site Inspection.
-Geophysical methods.
-Quality of records for
 penetration.
-Seismic data may not
 resolve small-scale
 features.
-Geologic naps/cross sections
 and ran data for seismic data.
-Reports on penetrations/back-
 filling activities should be
 provided.
504    DISPOSAL/STORAGE

-------
                                                                 Table 3
                                         Hydrologic Information for Repository Performance Analysis
Property
Surface
Mater
Information Needs
-Regional and local
water budgets.
-Hydraulic network.
-General water chemistry
and use.
Characterization Methods
-Site field survey.
-Literature searches for
soil, meteorological, and
water quality data.
-Soil boring and well logs
for water table elevation
determination.
Limitations to the
Characterization
-Problems with determining
area extent of field
study.
Presentation of Data
-Information presented In text
form plus:
o surface topographic and
hydrologlc maps of 1:34,000
scale
o SCS soil maps
o tabulated surface water flow,
volume, and quality data.
           Ground
           Mater
-Ground-water location.
-Aquifer characteristics.
-Ground-water chemistry.
-Geophysical surveys       -Geophysical resolution.
-Field ground-water testing -Location of well  boreholes
 and sampling.             and rock coring activities.
-Laboratory analyses.       -Inaccuracy of ground-water
                        flow models.
                             Table 4
                   Design and Operation in Mines
        Factor
                                       Relevant Information
Structural Stability of Design
Waste Emplacement
Waste Characteristics and
Compatibility
Mine Closure Seals
         Rock strength and stability.
         Size, depth, and spacing of underground
         openings.
         Equipment emplacement procedures used
         to convey wastes.
         Emplacement and backfilling methods.
         Maste reactivity and volatility.
         Waste compatabllity with engineered
         structures.
         Solubility of waste In ground waters.
         Oegradability of waste components and
         reagents 1n  repository conditions.

         Strength and chemical longevity.
         Location and design load of seals.
 PERFORMANCE APPROACH
   The performance approach to permitting a waste repository must
 provide the following types of information to be demonstrable:

 •  Steady-State and transient natural conditions
 •  Conceptual model of system behavior
 •  Performance allocation for system based on release criteria
 •  Component performance specification
 •  Methodology for detailed component and global system response
   modeling
 •  Interfacing detailed component and global system models

   This information will enable the permitting agencies to evaluate
 both  natural and man-induced scenarios  and system responses
 separately and provide the permit applicant the necessary tools to
 assess engineering and economic  feasibility.
   To develop these system and component models and understand
 system behavior, insitu geology, hydrogeology, geochemistry and
 geomechanics of the steady-state natural formations must be con-
 ceptually analyzed  using the data  cited  earlier. This analysis
 provides a baseline from which to gauge how mine construction
 has influenced the natural groundwater flow system and mechani-
 cal properties of the host. In all likelihood,  mining operations have
 been  conducted  and  shaft seals  designed and  constructed to
 optimize economic return with little concern for long-term perfor-
 mance, thus inducing potentially unfavorable transient conditions.
These transient conditions are a critical aspect of understanding
 system response upon repository closure and return to steady-state
regional  conditions.
  Once the steady-state and transient site conditions  have been
established, a conceptual model of  the waste repository  system
-Hydrologic setting information.
-Geophysical data should be
 provided.
-Stratigraphic cross section.
-Tabulated chemical data.
-.Hydraulic conductivities and
 transmlssivities specified.
-Geochemical speciation/
 saturation maps.
                                               behavior can be developed. A repository's isolation configuration
                                               will depend on the stability and permeability of the host rock and
                                               overlying formations and water table location. Figure 3 shows a
                                               schematic conceptual design of one type of repository.

                                                                         .	SURFACE PLUG
                                                                         SHAFT
                                                         BULKHEAD
                                                                           Figure 3
                                                        Schematic Illustrations of an Isolation Configuration

                                                 This conceptual model provides a reference point for the most
                                               complex  and geologically expected  case  for salt and limestone
                                               mines. Here, the repository horizon is dry despite an overlying
                                               water table that is separated by an aquitard. To maximize waste
                                               isolation, mined rooms and drifts  would be backfilled and panel
                                               bulkheads emplaced. The shaft would be backfilled at least up to
                                               the  uppermost bulkhead and a surface plug  emplaced. Upon
                                               closure, a variety of processes will occur and the conceptual per-
                                               formance model must account for them. For instance,  a model
                                               is needed to assess the hydrologic response of the bulkhead region
                                               at the water table when backfill consolidates. In this case, the model
                                               will need to specify repository geometry, individual components,
                                              .material parameter ranges, processes and boundary conditions. The
                                                                                                    DISPOSAL/STORAGE     505

-------
Likelihood of increasing flow around a bulkhead when backfill con-
solidates is high; therefore, flow barrier backups would be criti-
cal.  Since system-breaching scenarios are of primary concern,
component  performance allocations need to be established for
engineering- and  construction-related aspects of the  repository
based on scenario occurrences.
   Performance allocations for system components are needed
before component performance specifications can be defined. In
this  step, decisions on what components will be needed to satisfy
the overall system release criteria must be identified. For instance,
in a  stratigraphic sequence, determining how many bulkheads are
needed to prevent a barrier and how much water flows through
the section to the repository horizon. By allocating a measure of
performance to components of the overall system, each compo-
nent's performance specification can be defined for post-closure
design and construction. The information collected and assessed
will  be needed to conduct both global and component response
calculations.
   A two-fold strategy to predict global and component response
must be implemented to predict post-closure performance. One
effort evaluates the detailed performance of repository components
such as the host rock, waste containers and forms, liners, seals and
bulkheads.  The other effort evaluates total system performance
using coupled simplified numerical functions derived from the more
complex component models. The global models are used to evalu-
ate the hydrologic response  of all the components. This metho-
dology assumes  that  complex component  behavior such  as
host-rock creep, for instance, is evaluated in sufficient  detail that
the  system  hydrogeologic  response and, hence, waste release
behavior can be predicted. To evaluate component behavior, a
methodology that comprises calculations of  both chemical and
physical constitutive behaviors as a function of time  is needed.
These constitutive behaviors may have a theoretical or empirical
basis and may require validation. The numerical models used to
represent the conceptual model must be able to represent the
behavior of a complex engineered  system in  the environment.
  As an example, sealing system response to groundwater flow is
critical. A number of component models will be needed to assess
detailed response of the following components to hydrological, geo-
chemical, mechanical and possibly thermal (i.e., if biologic waste
degradation yielding heat occurs) effects:

• Host rock and  disturbed zone
• Backfilling materials
• Waste containers
• Cementitious bulkheads and liners
• Penetrated overlying strata and associated  disturbed zones

  Each  component will undergo a variety of natural processes that
influence key parameters. In this case, the parameters  of impor-
tance to assess  hydrologic  response are  material and interface
permeability, component strength and component longevity. The
processes shown in Table 5 must be evaluated for  this case.

                           Table 5
          Key Processes Influencing Component Response
  Penetrated Rock and Disturbed Zone

  Waste Containers

  Backfill


  Bulkheads
       Process

Creep, Dissolution, Healing

Corrosion, Dissolution

Consolidation, Dissolution,
Interface Bonding

Dissolution,  Interface Bonding
                               for a sealing system. Table 6 is a partial list of those codes and
                               how they can be used to predict component response as a function
                               of various processes.

                                                           Tabk 6
                                       Partial List of Existing Code* to Simulate Key Processes
                                       Affecting  Component Response and Groundwater Flow


                                      Code	ter Process to be S1ปlate4
                               STCAIM/CAVS (6)

                               Ed3/6 (7)

                               SUTRA (8)

                               CMM (9)

                               DYปA-FLflW (10)
                                    Deforซtlon. Healing, and Creep

                                    Dissolution and Corrosion

                                    Sซturat*d-Unsaturซttd Flo.

                                    Saturated Flw

                                    Consolidation
  Each process can be simulated and predicted to help understand
repository component performance response. A number of existing
computer codes are available to help predict long-term response

506     DISPOSAL/STORAGE
  The interrelationships of the processes to be modeled are exten-
sive, yet the resulting code output for this case must be in terms
of permeability, strength and longevity. The output can then serve
as input to the global model to simulate how  the total system will
respond as a result of natural processes affecting the sealing com-
ponents.
  A growing  number of codes can simulate global  response
processes  for repositories, most of which come  from high-level
waste projects. One code that  can be used to assess post-closure
global hydrologic response among other aspects of a total system
performance assessment is SYVAC5.  SYVAC is unique in several
areas but what makes it attractive for analyses of repository' per-
formance  is that it:

• Provides overall system approach to enable comparative studies
  of component effectiveness
• Incorporates variable uncertainties
• Produces estimates of impact distributions statistically
• Uses simplified descriptions of all processes that can be estimated
  during component modeling

  It is the output from SYVAC or another similar code that will
provide an ability to predict  system hydrologic response as a
function of component design and important natural processes.
By combining results from the more  detailed component studies
with a global systems  analysis, a permit applicant  will  have a
demonstrable performance approach for licensing a geologic reposi-
tory for hazardous wastes.

CONCLUSIONS
  Geologic repositories for untreated CERCLA and RCRA wastes
are a viable option for  long-term isolation from the environment
of which salt, limestone, potash and fluorspar are the most attrac-
tive hosts. To predict repository performance requires extensive
site characterization data and a method to evaluate how long-term
processes will influence the post-closure hydrologic isolation of the
system. A two-fold strategy is presented as  a model on  how to
predict global and component responses once the  repository is
closed. Using this approach, detailed assessments of repository
component behavior due to natural phenomena  such as dissolu-
tion, consolidation,  creep and other causes  for  the components
present such as host rock,  waste containers and  seals  can  be
predicted  using a variety of existing computer codes. The output
of these models can then be simplified  and serve as input to a global
systems response simulation. The output from the global systems
simulation will provide a permit applicant with a defensible estimate
of whether or not a  proposed repository will perform such that
a "no-migration" permit  variance can be established.
REFERENCES
1. National Academy of Sciences, Board  on Radioactive Waste Manage-
   ment, A Study of the Isolation System for Geologic Disposal of Radio-
  active Wastes, National Academy Press, Washington, DC, 1983.

-------
2.  The Earth Technology Corporation, "Evaluation of the Feasibility of
   Hazardous Waste Disposal in Geologic Repositories," Final  Draft
   Report, prepared under U.S. EPA Contract No. 68-01-7310,  Work
   Assignment 11, 1986.

3. The Earth Technology Corporation, "Guidance Manual for Hazardous
   Waste Disposal  in Geologic Repositories," Draft Report, prepared
   under U.S. EPA Contract No. 68-01-7310, Work Assignment 37, 1987.

4. Ertec, The Earth Technology Corporation,  "Information Needs for
   Characterization of High-Level Waste Repository Sites in Six Geologic
   Media," NUREG/CR-2663, 1985.

5.  Goodwin, B.W., "The SYVAC Approach for Long-Term Environmen-
   tal Assessments," Preprint of the Symposium on Ground-Water Flow
   and Transport Modelling for Performance Assessment of Deep Geo-
   logic Disposal of Radioactive Waste and Critical Evaluation of the State
   of the Art, Alburquerque,  NM, 1985.
6.  Hofmann, R., "STEALTH—A La Grange Explicit Finite Difference
   Code for Solid Structures and Thermohydraulic Analysis, Introduction
   and Guide,"  Volume 1, ERRI NP-2080-CCM-SY,  1981.
7.  Wolery, T.J., "Calculation of Chemical Equilibrium Between Aque-
   ous Solution  and Minerals: The EQ3/6 Software Package," UCRL-
   2658, Lawrence Livermore National Laboratory, 1979.
8.  Voss, C.I., "SUTRA, A Finite-Element Simulation Model for Saturated-
   Unsaturated,  Fluid-Density Dependent Groundwater Flow with Energy
   Transport or Chemically Reactive Single Species Solute Transport,"
   USGS National Center, 1984.
9.  The Earth  Technology Corporation, "Documentation of a Three-
   Dimensional Flow Model—GRAM," prepared for Office of Nuclear
   Waste Isolation, Battelle Memorial Institute,  Columbus, OH, 1984.
10. Prevost, J.H., "DYNA-FLOW: A Nonlinear Transient Finite Element
   Analysis Program," Dept. of Civil Engineering, Princeton University,
   1981.
                                                                                                       DISPOSAL/STORAGE     507

-------
                 Numerical  Evaluation System  for  Comparison  of
                                   Potential  Land  Disposal Sites

                                                     Gwen S. Ruta
                                      U.S. Environmental  Protection Agency
                                                        Region I
                                                Boston, Massachusetts
                                                 Geoffrey W. Watkin
                                           Jacobs  Engineering Group Inc.
                                                 Martinez, California
ABSTRACT
  Local land disposal is one remedial alternative under consider-
ation for cleanup of river sediments contaminated with PCBs in
Massachusetts. A quantitative landfill site evaluation system was
developed to assist regulators and community members in deter-
mining the feasibility of this option and identifying acceptable and
appropriate land disposal sites for further consideration. The
system, designed to  give a comparative ranking of each site in
relation to the others, requires only information which is readily
obtainable from published maps and reports.
  The system was used to rank 33 potential land disposal sites in
the project area. Out of a maximum possible score of 198, the site
scores ranged from 79 to 146 and were well distributed. Based on
this evaluation of potential sites, a local land disposal facility on
which to dispose contaminated  sediments appears feasible.

INTRODUCTION
  The  sediments of  the Housatonic River, which runs through
western Massachusetts and Connecticut to Long Island Sound, are
contaminated with over 35,000  pounds of  PCBs.  The States of
Massachusetts and Connecticut have  issued health warnings against
consumption of fish and frogs for over 100 river miles because of
the presence of PCBs in concentrations over the U.S. Food and
Drug Administration's acceptable level. The presence of the PCBs
is primarily due to the past operations of a large transformer
manufacturing facility located in the Berkshire region of western
Massachusetts. Over 90% of the PCBs are contained in approxi-
mately 250,000 yd3 of sediments in  a 12-mile stretch of the river
between the facility and the first downstream  impoundment1.
  Several options are under consideration for remediation of the
river system, including dredging and  local land disposal of the most
contaminated sediments. In order to evaluate the feasibility of this
option given the hydrogeology, population distribution, environ-
mental characteristics, land use patterns and  regulatory constraints
of the area, a quantitative site evaluation system was developed.
The system was designed to determine the feasibility of pursuing
any remediation technique which required local land disposal and
to simultaneously screen potential sites as land disposal options.
As a screening technique, the system relies  on  data sources such
as federal, state and local maps and  reports, academic studies and
regulatory requirements and criteria.
  The site evaluation system is  neither designed to result in the
selection of a site nor to make determinations about whether poten-
tial sites meet all regulatory criteria. The evaluations will be used
simply to determine the feasibility of local land disposal remedial
alternatives; the site scores can be used only on a comparative basis
and do not relate to any absolute score for "good" or "bad" sites.
Should local land disposal be selected as the remedial alternative,
the information developed for the site evaluation system could be
used as a basis for a site selection and approval process involving
federal,  slate and local  regulatory agencies  and  the affected
communities.
LAND DISPOSAL OBJECTIVES
  The first step in developing the site evaluation system was to
identify a set of land disposal objectives. The evaluation system
would then measure the degree to which proposed sites met those
objectives. After an extensive literature search and several oppor-
tunities for local community input, the following objectives were
developed for the selected site:

  Protect community health
  Ensure the longterm structural stability of the disposal unit
  Protect groundwater quality
  Protect air quality
  Protect ecologically sensitive areas
  Be consistent with community plans and goals
  Be consistent with the state and regional general plan
FACTORS
  Numerous factors affect the degree to which any specific site
meets the land disposal objectives identified above. For ease of
discussion and analysis, the factors identified for the site evalua-
tion system were grouped into five general categories:

  Engineering and geotechnical
  Hydrogeological
  Public health and community impact
  Environmental and ecological
  Socioeconomic

  For each factor, specific evaluation criteria were developed which
reflect the degree to which a potential site meets the land disposal
objectives. The criteria were derived from many sources  including
local, state and  federal regulations, current literature on land
disposal facility siting2"1, current technical literature in each of the
factor categories, case  studies addressing similar problems and
sound environmental and engineering judgement.
  The criteria for each factor are used to determine a factor score
for the site which ranges from 0 to 3. The scores are defined as
follows:
 508    DISPOSAL/STORAGE

-------
Score                            Definition

  0         Unacceptable—The potential site does not meet evalu-
             ation criteria defined by regulations or required to
             meet land disposal objectives.
  1         Marginal—The potential site does not meet the criteria
             defined by regulations but may meet the criteria re-
             quired to meet land disposal objectives.
  2         Acceptable—The  potential  site  meets  the criteria
             defined by regulations or required to meet land dis-
             posal objectives.
  3         Exceptional—The potential  site exceeds the criteria
             defined  by regulations or  required to meet  land
             disposal objectives and provides an identifiable benefit
             in reducing risk to public health and the environment.

Engineering and Geotechnical Factors
  The factors and criteria for this category are shown in Table 1.
These factors are directly related to the engineering feasibility of
building a secure land disposal facility at each site.  Sources such
as U.S. Geological Survey (USGS) topographic, soil and seismicity
maps; National Flood Insurance Program maps; U.S. Army Corps
of Engineers reports; county sewer system and road maps and utility
company maps were  used to gather the information necessary to
rate  the sites.  The determination of site values for the factor
"estimated site capacity"  was not conducted within the scope of
this project, but would be  based on available topographic infor-
mation, available acreage at the site, depth to groundwater, depth
to  bedrock, design requirements and sound engineering judgment.
                              Table 1
                              Table 2
                        Hydrological Factors
  pjctor
  pemnblllty of SurflctaY N*t*rlal.
   IK-hydraulle conductivity, oVeec)
  Soil 4tteiuatlon
                        _0_

                       K>10~*
                       0-J5>
                       CUy I allt
          15-501
          cl*y I lilt
                                           n>Kr*
50-75%
eUy i silt
75-100*
Clay t tilt
Depth to Historic High water
Table
in-plaee Sell Thickness
Prov.ii.ty to Existing,
Plamd or Potential
Underground Souroee
of Drinking Mater
Destination oC Site Drainage
<4 (t
<2 (t
Claas I
gnundwatar
water
Class A
surface water
>'"
>2ft
Class II
groundwater
water
Potential
Claas A
surface water
ซo ft
I* ft
no drinking
water source
within 1000 ft
Not Class A
surface water
>100 (t
>4 (t
no rtrinttlng
water source
within 1000 t
or else* III
groundwater
Not Class A
or t surface
water
  Site ftecharoe Characterletlai
                       Site In critical Recharge     Ho recharge to  tot In recharge
                       •qulCer      supplies current useable aojjlfer area
                       protection area  or potential
                                 sources or
                                 drinking weter
to quantify, the factors developed measure the suitability of each
site in relation to its surrounding community. USGS topographic
maps, county transportation reports, Massachusetts Department
of Public Works land use maps and county cultural and historical
landmarks guides were used to develop information on each site.


                             Table 3
            Public Health and Community Impact Factors
                                                              ฃ.ป..
 Hearby Population
                                                                           Current Land Ownenhlp and Use
                                           Ho houMa,
                                           workers within
                                           2000 ft

                                           negligible risk
                                                                                                                              accident
                                                                                                                              potential
                                                                                                Public preserve  Multiple owners,
                                                                                                          intensive uses,
                                                                                                          major existing
Engineering and Geotechnical
Rating
flood SuseeptablUty 1-100 yr 100-500 yr

Rtli

Are.

Sub




itlonshlp to Active Faults

•a of Rapid Geologic change

ildence Hazards



Liquefaction Potential


Dam

Loc



Failure Inundation Areas

at ion oE Permlttable Discharge


In actlva fault
zone
unstable, elope

Hi nod or
pumpod eras


High water table,
loose MnJu


Silo below dam

No direct
discharge, eewer
discharge H ml

<200 ft
setback
Unstable, level

Cubs Ide nee
exhibited, no
mining or pumping

High water table,
sands can be
dens if led

Ml thin Inun-
dation sons
No direct
discharge, sewer
discharge (1 mi
Factors
500 yr

>2oo et
setback
Stable, slopo

Stable, pumping
nearby


Deep water table
with loos*
sands

Near dam,
outside lone
Direct discharge
discharge (1 ml
JL SIC.
>SOO yr J

>3000 Et 3
setback
Stable, level 2

Stable, no 1
pumping or
mining nearby

Deep water table 2
with dense sands
silta or clays

Ho dam 2
nsarby
Direct discharge 1
or sewer
discharge \
corrjested

Significant
construction,
hiring needed

High, or
resources on.lt

Urge- scale
anomaly



Ecological
DWH in Tab)
Commercial,
light industrial,
open apace
Moderately
Incom latent
Up to 25%
congested

Slightly over
capacities


Moderate
e potential impact

Moderate scenic
Lrpact



Factors
le 4. identif
Heavy
, Industrial
Consistent
Ml thin
capacities

Within
capacities


Low potential
Upset

Law scenic
Inpact, few
viewpoints



v soecific
Mining, waste
treabnent/
disposal
Enhances
lirtplmentet Ion
Hell within
capacities

Mil within
capacities


Low potent 1 si
Impact, none
within 1 nl
Ho scenic
Impact, low
visibility



environir
  Availability of Adequate soils
Estimated Site Capacity
Surface Water Control
Access Roads
<100K cu yds
Excessive
runoff,
topo limited
tlo road, steep
dirt roads
100K-2SOK cu yds
Requires
extensive
grading
Unimproved road,
road under
conat ruction
250K-100K cu yds >3QOK cu yds
Moderate Little
si to grading
grading required
25' Width or 25' Width or
less, light more, stable
duty paved road paved road
  Site Utilities


  Buffer Zone
tal or ecological characteristics of each site. These characteristics
reflect the degree to which a site impacts its surrounding ecosystem
or is contained in an area of critical environmental concern. Again,
USGS topographic maps were used to develop information on the
sites. Endangered Species lists and Massachusetts National Heritage
Program reports were used.

                             Table 4
               Environmental and Ecological Factors
Hydrogeological Factors
  The factors and criteria for this category,  shown in Table 2,
reflect the degree to which a site has natural characteristics which
lessen the potential for migration of PCBs from the site or minimize
the impacts of such migration. Sources of information used to
develop scores for these factors include USGS soils and ground-
water contour maps,  county  water management plans, Massa-
chusetts surface water quality classifications and town  Planning
Board maps.
Public Health and Community Impact Factors
  Table 3 shows the factors and criteria in this category.  Two
primary concerns in siting a suitable disposal facility are the poten-
tial exposure of human populations to the contaminated material
to be disposed and the political opposition that may result from
perceived  risk by a nearby population. Although often difficult
 •ralmlty to Wetlands
 rlor*/Faun*i pn**nc* of
   •rd cntangtnd Sp*etM
 rLoปi rrMno
  QuUundlng
  r*iiwi HAblut c***ntUL fee
   Lit* 5t*o* or I* Part I* •
 Otullty of ll*tuปt Ch*r*ct*rlatlc
                                           outild* 100 Ct   ffetUnd. *n
                                           buffer ton*    .on then 1 nl
                                                     •1 1 Ml M*y
Preeent onelte  pment tn ซrM, praeent In ซnซ, AbMnt fro* en*  I
          llnlud l*pect  no l*Kwct

PrMent on.lt*  pceMnt In *r**, Pment In *ree, AbMnt frai ena  I
          Halted licxct  no livact

Contain*      Habitat or    Habitat or    DOM not contain  1
naoMeery     flyMy prMent  flywey preMnt  hitbitat, not part
habitat, part   In ana, lUltel In ana, no    at flyoay
of flyxay     l^>act      Impact

High oปallty   High guallty   No high quality Ho high quality  1
rercurcM,    reaouroM, eon. rMouroM,    revajroae, hMvy
undleturbatl    illaturhenoa    prevlou. UM   UM
Socioeconomic Factors
  By far the most difficult to quantify, the socioeconomic factors
and criteria shown in Table 5 were included to measure the extent
                                                                                                          DISPOSAL/STORAGE     509

-------
 to which a potential site would impact the regional economy. In
 some cases, these factors could not be quantified using readily avail-
 able information. In order to develop data for evaluation of these
 factors, surveys of the local labor supply, business community and
 real estate market would be required. Sources of information which
 were helpful in evaluating these factors included county tourism
 and historical landmark guides and case studies of other hazardous
 waste land disposal facility siting projects'-7.

                            Table 5
                     Socioeconomlc Factors
                                    KJUflg
  r-cm                  _t_        j_       _l_        _L    USป
                               DlKปmlblซ    InatgnlfttiMH
                                   MM,   OutlldB tCMTIl
                                                 Ok.'*ldi UBfrlll  1
  M9lwwl tncow ซA] Bvlo^MK
 SIGNIFICANCE VALUES
   In order to incorporate the relative importance of each factor
 into the overall site score, each factor was assigned a significance
 value ranging from  1 to 3. The significance values are based on
 the degree to which  the factor is directly related to fulfillment of
 the land disposal objectives and the ease with which any potential
 problems associated  with that factor can be mitigated. Significance
 values are defined as follows:

 Value                              Definition

   1                  Low importance—the factor only
                     indirectly affects the fulfillment of land
                     disposal objectives and can  be easily
                     mitigated.
   2                  Medium  importance—the factor is
                     directly related to  fulfillment of land
                     disposal objectives and can  be mitigated
                     without extensive effort.
   3                  High importance—the factor is directly
                     related to fulfillment of land disposal ob-
                     jectives and would be impossible  or
                     exceedingly difficult to mitigate.
                       The  significance values assigned to
                     each factor are shown on Tables  1 to 5.
SCORING
  The total score for each potential site is calculated by multiplying
the factor rating times the significance value for each factor to come
up with a sub-score for each factor and  then summing the sub-
scores for all the factors to obtain the total site score. To illustrate
the scoring process, an example is presented.

Example
  For the factor "Proximity to Wetlands," the factor ratings are
based on the Massachusetts Hazardous Waste Management Regu-
lations which state that no active portion of a landfill shall be con-
structed or expanded into wetlands. In addition, the Massachusetts
Wetlands Protection  Regulations establish a 100 ft wide area
bordering a wetland as a buffer zone. Therefore, the criteria for
this factor are defined as follows:

  0—Site is within  a  wetland
   1—Site is within  100 ft buffer zone
  2—Site is outside of 100 ft buffer  zone, abutting a wetland
3—Site is far from a wetland (greater than 1 mile)

  The significance  value assigned to this factor is 3, because the
proximity of the  potential  site to a wetland is directly related to
fulfillment of land  disposal objectives (protect ecologically sensi-
tive areas) and is impossible to mitigate (i.e., the wetland cannot
be moved).
  If the potential site were located outside the 100 ft buffer zone
for a wetland, its factor rating would be 2. The subscore for this
factor would then be:

  factor rating  x significance value = subscore
          (2)          x          (3)          (6)


PUBLIC  PARTICIPATION
  One reason for undertaking this project was that previous private
efforts to determine the feasibility of local disposal had met with
skepticism in  the local community.  Citizens believed  that the
previous studies were biased and inadequate. It was, therefore,
imperative that public input be sought in the early stages of this
project to ensure that community concerns were addressed and that
any conclusions  reached would  be received with confidence.
  Prior to beginning any substantive work on the site evaluation
system, a series of small meetings was held with members of the
community, the  media and local government officials to solicit
input on  the project. Comments  were received that addressed
factors to be considered in the site evaluation and their relative
importance to the community. Also, sites were suggested for evalu-
ation and visited  with their proponent if possible. These comments
and suggestions  formed the basis for the site evaluation system.
  After the draft evaluation system had been developed, a summary
of the system was mailed to concerned citizens and officials and
a public meeting  was held to take comments. Several valuable sug-
gestions were made which  helped tailor the system to the needs
and concerns of the local community.  For example, factors relating
to tourism and recreational land use, both major contributors to
the  regional economy, were given  high significance values.
  In addition, the county planning board was subcontracted to
provide local information  such  as traffic,  land  use  and zoning
patterns and to act as a liaison to the public. This helped to ensure
that accurate local information was obtained and also helped to
instill confidence in the project  in the local community.

RESULTS AND CONCLUSIONS
  The site evaluation system developed was used to rank 33 poten-
tial  land  disposal sites in the defined project  area. Out of a
maximum possible score of 198, the site scores ranged from 79 to
146 and were widely distributed within this range. Site scores were
analyzed at each significance level and in the five factor categories.
  Overall, site scores were higher on the least important factors
than on the most important. Also, better scores were seen  in the
engineering and geotechnical,  environmental and ecological and
socioeconomic categories  than in the hydrogeological and public
health and community impacts categories. Some sites, however,
scored consistently high over  all factor categories and at all sig-
nificance levels.
  Although the  Housatonic River  Valley is not ideally suited for
a PCB landfill facility, the use of the site evaluation system shows
that siting of such a facility in the project area is feasible. However,
a formal site selection process involving federal, state and local
agencies, the responsible party and the local communities would
be required  to implement  land disposal as a pan of any remedial
action. For any potential site to become a facility, the support and
approval of the local community or communities is required. The
fact that the potentially affected communities participated in the
identification and evaluation of potential sites contributes to the
potential for success  in siting a local facility.
  The site evaluation system described in this article can be easily
tailored for use in other areas by considering special hydrogeolog-
ic conditions, state or local regulations or community concerns.
Also, at a further stage in the siting process, the system may be
used to compare a small  number  of candidate sites by utilizing
detailed,  site-specific information to rate each factor.
 510    DISPOSAL/STORAGE

-------
DISCLAIMER

  The Site Evaluation System described above is not finalized. The
opinions or assertions contained herein are the writers' and are
not to be construed as reflecting the official views of the U.S. EPA.

REFERENCES

1. Stewart Laboratories, Inc., "Housatonic River Study, 1980 and 1982
   Investigations," Stewart Laboratories, Inc., Knoxville, TN, Dec., 1982.
2. Cartwright, K., "Selection of Waste Disposal Sites," Bulletin of the
   Assoc. Eng. Geol., XIX,  197-201.
3. Pulford, R.B. and Shilepsky, A.P., "Minnesota's Hazardous Waste
   Siting Criteria," J of Water Res., Planning Management,  109, 1983,
   165-173.
4.  Louis Berger & Associates, Inc., "Hazardous Waste Facilities Siting
   Manual," prepared for the Southern California  Hazardous Waste
   Management Project, Southern California Assn. of Governments, Los
   Angeles, CA, Jan., 1985.
5.  Applied  Economic Research, "Economic Impact Analysis: Stablex
   Hazardous Waste Treatment Facility, Hookset, NH," Applied Economic
   Research, Laconia, NH, 1983.
6.  U.S. Department  of  Energy, "Alternate Site  Selection Process  for
   UMTRA Project  Sites,"  Uranium Mill Tailings Remedial Action
   Project,  Albuquerque Operations Office, Albuquerque, NM, Mar.,
   1986.
7.  Clark-McGlennon Associates, "Proposed Solvent Reprocessing Facility:
   Socioeconomic  Impact Report," prepared  for SOLV, Inc., Clark,
   McGlennon Associates, Boston, MA, 1982.
                                                                                                    DISPOSAL/STORAGE     511

-------
                    Proposed  Short-Term  Burial/Storage  Method
                              For Unassayed  Hazardous  Waste
                                                Arthur  G.  Clem,  PE
                                              Clem International Ltd.
                                            Anderson,  South Carolina
ABSTRACT
  It is proposed that unknown and unassayed containers suspected
of containing noxious or hazardous waste may be stored tempor-
arily in a trench cut in the earth, encapsulated within an inorganic
mastic prepared from a sodium bentonite slurry and Portland
cement. The mastic is impermeable, K = 1  x  10 '7 cm/sec, and
easily  prepared and removed to recover the containers. A modifi-
cation of this method may be suitable for the burial of low level
nuclear waste.
INTRODUCTION
  Hazardous waste pirates represent a major problem for engineers
concerned with pollution control. A typical scenario follows this
general pattern: the State Police, or concerned citizens, report a
dozen or more rusting, unmarked drums dumped in a ditch along-
side a road. There is a strong probability that the contents are toxic
or hazardous, otherwise they would have been disposed of properly.
The contents are unknown. The quality and durability of the drums
is unknown. The degree of toxicity of the contents is unknown.
The potential of the products in two or more containers being an-
tagonistic is unknown. The sensitivity of the contents to heat and
cold is unknown. The list of unknowns can continue until the con-
tainers are removed, stored, sampled,  analyzed and a decision is
reached as to how the waste is to be handled or stored permanently.
  In a major hazardous waste handling facility, the unknown and
unassayed drums of waste are stored on durable concrete slabs,
with controlled drains and sumps in the event of leakage. Within
a short period of time, the contents are  known and the disposal
mechanism can  be selected.
  The situation is significantly  different when no major waste
handling facility is nearby. The containers must be protected from
weathering and the earth must be protected  from leaking liquids
until final disposal. An  encapsulation or wrapping is needed to
protect unassayed containers from the elements, protect aggres-
sive wastes from mutual attack  and halt leakage into the earth.
Protection is short-term, until the containers can be assayed and
moved.
  Gelled cement can provide the necessary encapsulation. Gelled
cement consists of a slurry of Portland cement,  hydrated in a
sodium bentonite suspension; it is poured around drums or con-
tainers stored within the  ground. The gelled cement  slowly stiffens
(it does not set hard) into a mastic resembling soft native clay. The
slurry has low permeability plus the ability to adsorb  and hold some
metallic ions, some of which may be  radioactive.
GELLED CEMENT
  Gelled cement consists of three readily available ingredients:
water, sodium bentonite and type 1 Portland cement. The ben-
tonite is mixed into a slurry with the water, using recirculation by
a centrifugal pump, and aged for  full hydration. The dry cement
is mixed with the bentonite slurry just prior to placement.

Sodium Bentonite
  Bentonite  is the common name for impure montmorillonite
(smectite), a clay mineral that hydrates with five times or more
of its weight of water, expanding or "swelling" to many times its
dry volume. Sodium bentonite is a highly impermeable clay mineral:
k =  l  = 10 "9 cm/sec with decades of success as a soil sealant
for oxidation lagoons, ponds, dams and  reservoirs.
  Unlike ordinary native clay, sodium bentonite is a gellant. For
example: a 3% aqueous suspension of sodium bentonite has a con-
sistency barely thicker than water; a 6lro suspension has the con-
sistency of gravy;  a 9^b  suspension has the consistency of
mayonnaise; and a  12% suspension has the consistency of axle
grease. Solids will remain in suspension  in a sodium bentonite
slurry.
  A bentonite slurry in water is non-filterable, a property needed
in oil well drilling and in slurry trench cut-offs. A 6% suspension
of sodium bentonite poured into  the earth will develop permea-
bility of k = 1 x  10"* cm/sec within  minutes, then decrease to
k =  1  x 10~7 cm/sec within a few days. For comparison, k  =
1 x 10 "7 cm/sec is the impermeability of high quality compacted
clay seals.
  Sodium bentonite has another  distinct  attribute: it has cation
exchange ability and will trade its sodium ions for calcium, mag-
nesium, cesium and strontium ions. Sodium bentonite has an ion
exchange capacity of 75 to 100 meg/100 g, highest of all clays.
  Cation exchange ability is important because it allows a ben-
tonite to trade a sodium ion for a heavier metal ion, lock it solidly
in position and hold it immobilized in place. High cation exchange
capacity adds an adsorbent insurance barrier to some radioactive
wastes buried for long periods of time.
  Sodium bentonite is best known for its use in well drilling,
bonding  foundry sands  and metal  ore  pellets,  but its most
interesting application in civil engineering is a hydraulic barrier
in soil and rock via soil-bentonite admixtures and slurry trenches
with and without Portland cement.
  Sodium bentonite is not a rare mineral. It is found in natural
condition  in large deposits in Wyoming, Montana, California,
Central Canada, Durango State,  Mexico and Argentina. A ben-
tonite capable of being improved by soda additions (activatable
 512    DISPOSAL/STORAGE

-------
bentonite) is found worldwide. Total annual production worldwide
is more than  6 million metric tons.

Portland Cement
  Portland cement reaches its maximum strength when the cement
to water ration is 2:1 (2 kg Portland cement to 1 1. water). As the
cement ratio  is reduced, the strength reduces sharply. When the
cement ratio is ultra-low (e.g., 0.1:1) the water volume is so high
that the cement  settles out of suspension, settles on the bottom
of the container and hardens there. The uncombined water remains
on top. However, when the low ratio Portland cement (0.1:1) is
added  to a sodium bentonite suspension, the particles of cement
remain fully suspended while they hydrate. As a result, the initial
mixture of cement and sodium bentonite slurry is relatively low
in consistency but thickens later into a uniform soft mastic.
  The properties of the gelled cement are controlled by the propor-
tions of bentonite slurry and dry cement additions. For example,
if high impermeability is demanded and ultimate strength is of lesser
significance, a mixture of 100 kg of cement in 10001. of 6% sodi-
um bentonite slurry (0.1:1) should yield a permeability of  1  x
10~7 cm/sec,  but with a set strength of less  than 2 kg/cm2.  If
higher strengths are needed, the formula is raised to 160 kg of
cement in 10001. of 4% sodium bentonite slurry. The impermea-
bility changes to the range 1 x 10~6 cm/sec or 5 x  10~6 cm/sec.
The set strength would be about 5 kg/cm22.
  The sodium bentonite level determines the degree of impermea-
bility;  the cement level determines strength.
  Gelled cement, also known as cement-bentonite (C-B)  slurry, is
used in civil engineering to construct hydraulic cut-off walls below
grade. A trench is dug in the earth through a slurry which replaces
the excavated earth. The product solidifies into a mass with the
consistency, strength and impermeability of a native clay soil.
TEMPORARY BURIAL OF UNASSAYED WASTE
  The objectives in the safe short-term storage of unknown wastes
are safety, ease of construction, reliability of the encapsulation or
wrapping, ease of removal and safe disposal of the temporary
sealant. These are achieved by carefully burying the containers in
a trench, sealed in a mastic following the general outline given
below:

• First, a trench is excavated in the earth, large enough in size to
  hold the unassayed drums. The side walls should be as vertical
  as possible, with a slope of 0.25:1. The trench bottom must drain
  to prevent a "bathtub effect."
• Next, the bottom 0.5 m of trench should be filled with granular
  aggregate, sand or fine gravel to allow any leachate or precipi-
  tation to drain.
• Then a clay or bentonite  lined sump to collect and hold any
  leachate should be installed.
• Unassayed drums, on wood pallets, are placed into the trench.
  Sound drums may be stacked 2 pallets (about 2 m) high; ques-
  tionable drums should be only one pallet high.
• The stacked drums in the trench should be enclosed with
  temporary forms to contain the gelled cement.
• A well-hydrated 6 % sodium bentonite slurry is prepared by
  mixing 60 kg of bentonite with 1000 1 of water. This slurry can
  be made in advance.
• In a separate mixing vat, 100  kg of  dry Portland cement  are
  added to the 1000 1 of 6 % bentonite  slurry and mixed until
  smooth and creamy. Note: the mixture will be about pea soup
  consistency.
• The gelled cement slurry is pumped over the stacked drums in
  the trench until the drums are completely encased in gel.
• The complete pile is covered with a plastic film to halt evapora-
  tion of water from the gel.
• As a result of these steps, the containers are stored within a water
  tight mastic. Should a drum or container leak, the leakage will
  be confined to its immediate location.
  When drums are to be removed for final disposal, lift truck forks
should be inserted to raise the pallet and break it away from its
encapsulation. The drum can be assayed and disposal arranged.
The gelled cement coating will dry and shrink when exposed to
air and can be disposed of in a suitable landfill.

LOW LEVEL NUCLEAR WASTE
  Short-term burial has a significantly different connotation for
nuclear waste as compared to hazardous waste. Nuclear waste is
sorted into two general groupings: low level waste (half life of 50
years  maximum) and high  level waste (half-life greater than 50
years), usually without regard to intensity of radiation. The reason
is sound. Nuclear waste decays into an innocuous element; 10 "half-
lives" reducing the hazard to a safe level. Waste with a half-life
of 50  years, the maximum for low level waste, will be spent and
safe in about 500 years. Higher half-life elements will still be potent
10,000 to  100,000 years in the future. Comparison of 500  years
against 10,000 years assigns low level nuclear waste the classifi-
cation of  short-term storage or burial.
  The safe burial method for high level waste is well established;
its long-term decay time ranging from 10,000 years to forever. Most
popular is a 500-meter deep shaft blasted from a solid granite
mountain, with waste enclosed in copper containers buried in holes
drilled in tunnels leading off the main draft. For increased protec-
tion against damage or leakage, the copper containers are sur-
rounded by compressed bricks of sodium bentonite plus a sodium
bentonite-sand admixture backfill for completely filling tunnels.
The bentonite halts water infiltration. It will also adsorb  some
radioactive cations  in the event of an accidental fracture and
leakage.
  Low level waste with its more rapid decay is handled as a short-
term problem. A trench is excavated in a clay soil, fitted with a
sand drain connected to a sump. Waste in boxes, containers and
drums on wood pallets is unloaded and stacked neatly in the open
trench. As the trench is filled, dry, free flowing sand is bladed into
the filled trench  to fill voids between containers. A cap of com-
pacted clay prevents natural precipitation from entering the burial
site. The totally filled site is then closed for the maximum life ex-
pectancy of 500  years.
  Ideal site conditions for the short-term burial of low level nuclear
waste occurs in only a few areas. There is a need for a fabricated
short-term disposal site that can be built in many areas now being
rejected for reasons of inadequate burial soils.
  The proposal  for short-term burial of low level nuclear waste
is similar  to that for short-term burial of  unassayed hazardous
waste: burial in a shallow trench, encased in a gelled cement mas-
tic. There is only one major difference;  the bottom, sides, ends
and top of the trench are lined with a bentonite coated polypropy-
lene sheet.

Bentonite Coated Geofabric
  A bentonite coated geofabric consists of a sheet of woven or
non, woven polypropylene to which a uniform layer of sodium
bentonite  granules is adhered. The rolls are usually 4 m wide, 25
m long and hold 5 kg of sodium bentonite/m2. The coated fabric
is only slightly permeable: k =  1 x 10~9 cm/sec to k = 2  10~lฐ
cm/sec. The finished sheet is strong, non-biodegradable, non, toxic,
easy to cut and easy to install. Water impermeable seams are made
by facing two bentonite coated sides together and stapling with
an industrial stapler.
THE PROPOSED METHOD
  The  instructions for the method follow:

• Excavate a trench, sloped gently downhill in a selected area.
• Attach a bentonite coated geofabric to the exposed end, sides
  and  bottom.
• Install a granular aggregate drain in the bottom,  connected to
  an impermeable sump.
                                                                                                DISPOSAL/STORAGE     513

-------
• Load containers of low level nuclear waste, on pallets, into the
  trench. Note: you should record the exact location of each con-
  tainer into a permanent record.
• Drape a sheet of uncoated geofabric at intervals within the trench
  to act as a dam.
              Gel Cement
BentonjU Fabric
                           Figure 1
           Typical proposed low level nuclear waste trench.
• Prepare gelled cement from hydrated sodium bentonite and Port-
  land cement and pump or flow the slurry  into the trench.
• As the trench fills, cover the top of the gelled cement with a
  bentonite coated geofabric.
• Cover the filled trench with about  0.6 m of fill dirt followed
  by 0.3  m of top soil. Plant with shallow rooted shrubbery.

  Should any container rupture, its contents will be encased within
the impermeable mastic and further protected from leaching by
the bentonite coated sheet.  For details see Figure  1.
                               CONCLUSION
                                 This  paper discusses  a  method of encapsulating and storing
                               unknown waste containers within an inorganic mastic composed
                               of sodium bentonite slurry and Portland cement. The mastic is im-
                               permeable,  long  lived and  will adsorb  and hold some waste
                               products.
                                 A final important note: this is only a proposal; it is, however,
                               a proposal based on the known and published properties of sodium
                               bentonite, gel cement and bentonite coated geofabrics. No long-
                               term studies of specific toxic, noxious or hazardous wastes have
                               been made.
514     DISPOSAL/STORAGE

-------
                 Impact  of  the RCRA Land  Disposal  Restrictions
                               On  Superfund  Response  Actions

                                                   James Antizzo
                                                 John Cunningham
                                     U.S. Environmental Protection Agency
                                                 Washington, D.C.
                                                  Kathleen Hutson
                                                 Amelia Heffernan
                                          Booz, Allen & Hamilton Inc.
                                                Bethesda, Maryland
ABSTRACT
  On Oct. 21, 1976, Congress enacted RCRA to protect human
health and  the environment from the  improper disposal of
hazardous waste and to conserve material and energy resources.
Amendments to RCRA, enacted through the Hazardous and Solid
Waste Amendments (HSWA) of 1984, impose  substantial new
responsibilities on those who handle hazardous wastes. In parti-
cular, the amendments focus  on the land disposal of hazardous
wastes.
  Recognizing that  land disposal technology cannot guarantee
perpetual containment of all waste constituents,  Congress has
added provisions to HSWA designed to minimize reliance on land
disposal. Amendments to section 3004 of RCRA specifically pro-
hibit the continued land disposal of hazardous wastes beyond speci-
fied dates unless the wastes are treated to a level or by a method
specified by the U.S. EPA or are subject to an  exemption or
variance.
  Contaminated soil and debris from CERCLA 104 and 106
response actions and RCRA corrective actions are not subject to
the land disposal restriction provisions until November 1988 at the
earliest when some contaminated soil and debris may be restricted.
  This paper focuses on  the land disposal restrictions from a
Superfund perspective and on the U.S. EPA's on-going program
to develop land disposal restriction rules for contaminated soil and
debris from CERCLA 104 and 106 response actions and RCRA
corrective actions.

INTRODUCTION
  The following section provides highlights of the land disposal
restrictions and their applicability to contaminated soil and debris.

Hazardous and Solid Waste Amendments
  The Hazardous and Solid Waste Amendments of 1984 (HSWA),
enacted by Congress on Nov. 8, 1984, impose substantial new
responsibilities on those who handle hazardous wastes.
  In particular, the amendments  prohibit the continued land
disposal of untreated hazardous  wastes beyond specified dates,
"unless the Administrator determines that the prohibition...is not
required in order to protect human health and the environment
for as long as the wastes remain hazardous..." (RCRA sections
3004(d)(l), (e)(l), (g)(5), 42 U.S.C. 6924 (d)(l), (e)(l), (g)(5).
  Wastes treated in accordance with treatment standards set by
the U.S. EPA under section 3004(m) of RCRA are not subject to
the prohibitions and may  be land disposed. The statute requires
the U.S. EPA to set "levels or methods of treatment, if any, which
substantially diminish the toxicity  of the waste or substantially
reduce the likelihood of migration of hazardous constituents from
the waste so that short-term and long-term threats to human health
and the environment are minimized" (RCRA section 3004(m)(l).
42 U.S.C. 6924 (m)(l)).
  Land  disposal  prohibitions are effective immediately  upon
promulgation unless the U.S. EPA grants a nation-wide capacity
variance or a case-by-case extension.
  For the purposes of the land disposal restrictions program, the
legislation specifically defines land disposal to include, but not be
limited to, any placement of hazardous waste in a landfill, surface
impoundment, waste pile, injection well, land treatment facility,
salt dome, salt bed formation, underground mine or cave (RCRA
section 3004(k), 42 U. S.C.  6924(k))
  Congress also has prohibited the storage of any hazardous waste
that is subject to a prohibition from one or more methods of land
disposal unless "such storage is solely for the purpose of the
accumulation of such quantities of hazardous waste as are neces-
sary to facilitate proper recovery, treatment or disposal" (RCRA
section 3004(j), 42 U.S.C. 6924G)).
  The legislation  sets forth  a series of deadlines  for U.S. EPA
action. At certain deadlines, further land disposal of a particular
group of hazardous wastes is prohibited if the Agency has not set
treatment standards  under section 3004(m)  for such wastes or
determined, based on a case-specific petition, that there will be
no migration of hazardous constituents from the unit for as long
as the wastes remain hazardous. Other deadlines cause conditional
restrictions.
  On Nov. 7, 1986, the U.S. EPA promulgated the first set of land
disposal restriction (LDR) regulations. This  rule  established the
framework for implementing the LDR program and established
specific  treatment standards and  effective dates  for hazardous
wastes included in the first phase of the land disposal restrictions
(i.e., certain dioxin and solvent-containing hazardous wastes). On
July 8,1987, the Agency promulgated regulations  restricting land
disposal of hazardous wastes included in the second phase of the
land disposal restrictions (i.e., the "California list" wastes).

APPLICABILITY OF THE LAND DISPOSAL
RESTRICTIONS AT SUPERFUND SITES
  Hazardous wastes from Superfund  sites that  are to be land
disposed after the statutory effective dates are subject to the land
disposal restrictions  Section 3004(h) of HSWA exempts certain
wastes generated from CERCLA activities from the RCRA LDR
until Nov. 8, 1988. This exemption applies only to contaminated
soil and debris resulting from CERCLA Fund-financed actions
(section 104) and exercise of CERCLA enforcement  authority
                                                                                         DISPOSAL/STORAGE    515

-------
(section 106). The exemption applies only to contaminated soil and
debris which have been defined to include, but not be limited to,
soil, din and rock, and other natural and man-made materials such
as contaminated wood,  stumps, clothing,  equipment,  building
materials and storage containers. The exemption also may apply
to liquids or sludges that are mixed with soil or debris; applica-
bility of the exemption to such mixtures will be determined on a
case-by-case basis.
  The  exemption does not apply to  CERCLA bulk wastes  that
clearly are not contaminated soil or debris.  Nor  does the exemp-
tion apply to contaminated soil and debris that result from State-
ordered,  State-funded or private party-funded  responses taken
under the authority of CERCLA or exclusive of this authority.
Wastes not included in the exemption and prohibited from land
disposal are subject to the schedule and requirements imposed by
the LDR program. The effective dates are presented in Table 1.

                           Table 1
        Land Ban Effective Dales for Waste at Superfond Sites
     RCRA BULK WASTES


          SPENT SOLVENTS & DIOXIN-CONTAINING
          WASTES

             2 YR. NATIONAL VARIANCE FOR
             CERCLA WASTES (NON-SOIL & DEBRIS)

          SPECIRED WASTES (CALIF. LIST)

             2 YR. NATIONAL VARIANCE FOR
             PCBs AND HOCs

         AT LEAST 1/3 OF ALL LISTED
         HAZARDOUS WASTES

         AT LEAST 2/3 OF ALL LISTED
         HAZARDOUS WASTES

         ALL REMAINING LISTED HAZARDOUS AND
         CHARACTERISTIC HAZARDOUS WASTES
    CERCLA SOIL AND DEBRIS


         CERCLA SOIL AND DEBRIS FROM
         104 AND 106 RESPONSE ACTIONS
NOV. 8. 1986


(NOV. 8.1988)


JULY 8, 1987

(JULY. 8.1989)


AUG. 8,1988


JUNE 8, 1989


MAYS, 1990






NOV. 8.1988
     CONTAMINATED SOIL AND DEBRIS FROM STATE-ORDERED,
     STATE-FUNDED AND PRP RESPONSES MUST COMPLY WITH
     THE EFFECTIVE DATES FOR RCRA BULK WASTES
  DEVELOPMENT OF THE REGULATORY PROGRAM
FOR CERCLA SOIL AND DEBRIS
  The Agency is developing a program to regulate contaminated
soil and debris wastes that are covered by the exemption. The
Agency has several goals in developing this program. First, the regu-
lations must address the unique  aspects of CERCLA soil and
debris, including: physical and chemical characteristics of soil and
debris matrices and their effects on waste handling, pre-processing
and treatment; high degree of heterogeneity and variability within
and among waste  streams;  complex mixtures of wastes and con-
stituents; unknown sources of hazardous contaminants;  limited
ability to minimize wastes; inability to prevent generation of waste
through process modification; alteration of chemical and physi-
cal characteristics of waste matrices and contaminants through
weathering and other processes; the large volume of waste to be
addressed; the need to provide for the use of mobile treatment
facilities and equipment as well as fixed facilities to treat the waste;
and the relative lack of experience and data on handling and treat-
ment of soil and debris.
  Second, the Agency would like to provide flexibility in the regu-
lations to encourage treatment via a variety of treatment tech-
nologies including innovative technologies as they become available.
  Third, the Agency wants to ensure that the treatment standards
for soil and debris address potential cross-media impacts, thereby
ensuring that the BOAT standards indeed represent "Best" treat-
ment standards from both environmental and health perspectives.
  The Agency has developed and is implementing a regulatory
program to meet these goals. Major components of the program
include:
• Collection  and Analysis of Available Treatment Data—The
  Agency is conducting a comprehensive effort to collect all avail-
  able data regarding the handling, pre-processing and treatment
  of contaminated soil and debris. The data collection effort is
  being conducted in close cooperation with the Department of
  Defense and the Department of Energy as well as other Federal
  and state agencies and  non-government organizations.
• Research and Development  Studies—The U.S. EPA Office of
  Research and Development  is conducting a series of studies to
  assess the treatability of contaminated soils using a variety of
  treatment technologies, including: soil washing, incineration,
  low-temperature thermal desorption, KPEG and solidification.
  The studies are designed to assess the performance and limita-
  tions of the various technologies for a variety of wastes and to
  assess the sensitivity of technology performance to variations
  in waste characteristics and technology parameters.
  teristics and technology parameters.
• Field Studies—In addition to the above efforts, the Agency is
  conducting cooperative studies that are focused on  obtaining
  treatment data from large-scale, on-site treatment operations at
  Superfund  sites.
• Identification, Evaluation and  Resolution of Technical and
  Policy Issues—Development of an effective LDR regulatory
  program for Superfund soil and debris requires detailed evalu-
  ation and resolution of a wide variety of technical and policy
  issues. An Agency-wide work group with representatives from
  the Headquarters Offices, Regions and laboratories has been
  holding a series of meetings  to ensure that technical and policy
  issues of concern to various organizations are fully addressed
  in the regulatory program.
• Survey of CERCLA Soil and Debris—The Agency also is con-
  ducting a survey of all CERCLA 104 and 106 response actions
  to develop  a data base  on volumes and characteristics of soil
  and debris wastes. The data will be used to estimate the volumes
  and characteristics of wastes by treatability class and assess the
  adequacy of treatment  capacity.
• Evaluation of Treatment Types and Capacities—The Agency is
  evaluating treatment types available both on- and off-site. This
  survey is necessary  to  determine the  adequacy of  treatment
  capacity for each class of waste. The results of this evaluation
  and the survey of CERCLA  soil and debris will be used to
  determine whether the Agency should promulgate extensions to
  the effective dates of the LDR for contaminated soil and debris.

  The results of these Agency  studies and a variety of other
projects, including assessment of cross-media impacts and a Regula-
tory Impact Analysis, will provide the basis for developing the LDR
regulations for contaminated CERCLA soil and debris. The Agency
plans to publish a proposed Rule for selected CERCLA soil and
debris waste  in January 1988 and promulgate a Final Rule by
Nov. 7, 1988.

CONCLUSION
  The land disposal restrictions impose extensive limitations on
the land disposal of hazardous wastes. The U.S. EPA recognizes
that contaminated soil and debris waste pose unique handling and
treatment problems. Therefore, the U.S. EPA has initiated a variety
of studies, as discussed in this paper, that examine contaminated
soil and debris and provide data necessary for the promulgation
of a final rule by Nov. 7, 1988.
516    DISPOSAL/STORAGE

-------
                     Natural Resource Damages Under Superfund

                                           Richard W. Dunford,  Ph.D.
                                          William H. Desvousges, Ph.D.
                                            Research Triangle Institute
                                    Research Triangle Park,  North Carolina
SEMINAR OVERVIEW
  The Superfund Act extended the liability of businesses respon-
sible for oil spills or releases of hazardous substances to "damages
for injury to destruction of, or loss of natural resources." This
provision was recently implemented with regulations from the U.S.
Department of the Interior (DOI) defining the scope of such natural
resource damages and specifying the procedures for determining
them. These regulations are currently being challenged in court.
  The natural resource  damage provisions of the Superfund Act
and the accompanying DOI regulations  have  shifted some—
possibly, much—of the  cost of oil spills and hazardous substance
releases from the general public to the company responsible for
the release. We will examine these provisions and regulations in
our seminar, employing a very pragmatic approach. Audio-visual
materials will play an integral role in the seminar presentation. We
will illustrate each  component of a natural resource damage
assessment using a hypothetical example  of a hazardous substance
release into a major river. Key issues from a practical perspective
will be emphasized.

DETAILED SEMINAR OUTLINE
I.  Introduction

   A.  Relevant Clean Water Act  (CWA) and CERCLA/SARA
      provisions
      1. CWA provisions
      2. CERCLA provisions
      3. SARA provisions

   B. Development of U.S.  Department of the Interior regulations
      1. Draft  regulations
      2. Final regulations
      3. SARA amendments

   C.  Hypothetical release of hazardous substance into the Clear
      River (Table 1)
      1. Characteristics of the Clear River and its Basin
     2. Hypothetical hazardous substance release and associated
        natural resource injuries

II. Natural Resource Damage Assessment (NRDA) Process

   A. Overview of NRDA process (Fig. 1)

   B. Pre-assessment phase
      1. Purpose of pre-assessment phase
     2. Responsibility for pre-assessment phase
     3. Steps in pre-assessment  phase
  C. Assessment plan phase
     1. Purpose of assessment plan phase
     2. Responsibility for assessment plan  phase
     3. Steps in assessment plan phase
  D. Implementation of Type B Assessment
     1. Purpose of Type B assessment
     2. Responsibility for Type B assessment
     3. Steps in Type B assessment
  E. Post-assessment phase
     1. Purpose of post-assessment phase
     2. Responsibility for post-assessment
     3. Steps in post-assessment phase

III.  Phases in a Type B Assessment

  A. Overview of phases in a Type B assessment (Fig. 2)

  B. Injury determination phase
     1. Injury definitions
     2. Pathway determination
     3. Testing and sampling methods

  C. Quantification of effects phase
     1. Service reduction quantification
     2. Baseline services determination
     3. Resource recoverability analysis

  D. Damage determination phase
     1. Damages measured as lesser of restoration/replacement
        costs or diminution of use values
     2. Restoration/replacement costs
     a. Choice of restoration/replacement  methodology
     b. Estimating restoration/replacement costs
     3. Diminution of use values (Fig. 3)
     a. Preference of market-based valuation  methods
        over non-market valuation methods
     b. Exclusion of nonuse values' except when no use
        values can be determined

  4. Market-based valuation methods  for  estimating  lost use
     values
     a. Market price method
     b. Appraisal method (based on comparable sales ap-
        proach)
  5. Nonmarket valuation methods for estimating lost use values
     a. Factor income method
     b. Hedonic property value method
                                                                                                     SEMINARS    517

-------
                                   Table 1
          A  Hypothetical  Example of Natural Resource Damages
         To the Clear River From a Hazardous Substance Release

     Characteristics of  the Clear River and Us _gj_> in

              100 •lies  long, having an average width of one ปlle

              Classified (or  Drinking Water Supply,  fishing, And Swltvlng activi-
              ties by stale environmental  officials

          •    Many good pub He and private access points for recreation

          •    Supports extensive boating,  fishing, and svtmnlng activities
              throughout the  year

          •    The river basin covers several  thousand square lilies and contains *
              major Metropolitan arfซ (SOO.OOO people) about 25 Miles from the
              river

          •    Some substitute recreation sites arc available within 50-mile radius

     Hazardous Substance Release and Associated Natural Resource Injuries

              In the 1960s, ?50,000 gallons of NCT (a hypothetical petrochemical
              derivative) were plactd In a landfill,  which ts vlthln SO feet of a
              small tributary to the Clear River

          •    NCT Is toxic to fish and wildlife, and  ts a  known carcinogen

              For the last 10 years. NCI has  leached  fro*  the landfill into a
              major reach of  the Clear diver  (via the nearby tributary), eliminat-
              ing fishing and svlmmtng In  that reach

          •    Site cleanup took place tn 1967, which  stopped the NCT leaching and
              returned the niter quality In the river t>a- k to state standards

              Concentrations  of NCT are present In both fish and -Mdll'e ne*r the
              river
                                 Table 2
      Major Elements of the Natural Resource Damage Regulations
                 Being Contested in Court  by Petitioners

    legaj/constitut 'on_al_t_siues

        *    Restriction of damage recover/ to losses of  'public* uses

        •    No provisions  for pynlltvr damages

             Outage aปands BUS!  be used for restoration or replacement of  Injured
             natural resources

        *    Assessment  costs must be less than anticipated damage costs

        •    Restriction of scope of damages to nation or state ~t\ 4 vhoif
    Economic  issues
            Preference of marked-based valuation methods over nonmarkct  valua-
            tion methods

            Specification of  101 real discount  rate

            Damages as lesser of restoration/replacement costs or the diminution
            of  use values

            Preference of ->Illngnesi-to-pay criterion over willingness-to-
            accept criterion
                              Injury Determination
                            Quantification of Effects
                            Damage Determination
                                    Figure 2
                 Overview in Phases m a Type B Assessment
Martot-tMMd
nxxtnooi
Nonmwfert
nwthods
Ma/tart pnoe
AppraM (oompwaMa sales)
tntna
SknuMad
Other


Hadonlc property value
Travel CO*
ConOngem valuation
Una-day value
Other
                                 Figure 3
          Taxonomy of Valuation Methods for Measuring the
                        Diminution of Use Values
                            Pre-aปปMsmen< Screen
                              Aumament Plan
        Assessment
                                                        Assessment
                              Post Assessment
                                 Figure  1
       Overview of Natural  Resource Damage Assessment Process.
       c.  Travel cost method
       d.  Contingent valuation method
       e.  Unit-day  value  method
   6. Nonuse values
       a.  Option  values
       b.  Existence values
   7. Guidance on determining  damages
       a.  Effects  of remedial/response actions and natural
           recovery
       b.  Treatment of uncertainty
       c.  Discounting  past   and future  natural  resource
           damages
       d.  Role of substitute  resource uses
       e.  Scope of NRDA analysis

IV.  Judicial Challenges to Regulations

   A. Plaintiffs in lawsuits
       1.  States
   2. Environmental organizations
   3. Commercial firms/associations
518     SEMINARS

-------
  B. Major issues being contested (Table 2)
      1. Legal/constitutional issues
     2. Economic issues

V. Conclusions

  A. Brief summary
      1. NRDA phases
     2. Phases in a Type B assessment
     3. Judicial challenges to NRDA regulations
B. Implications of NRDA liability
   1.  Perspective of potentially responsible parties
   2.  Perspective of governmental trustees

C. Preliminary NRDA as an alternative to a formal NRDA

   1.  Purpose of a preliminary NRDA
   2.  Components of a preliminary NRDA
   3.  Advantages of a preliminary NRDA
   4.  Disadvantages of a preliminary NRDA
   5.  Conclusion
                                                                                                     SEMINARS    519

-------
        Contractor  Liability  and  Indemnification Under Superfund
                                            J. Kent Holland Jr., Esq.
                                        Wickwire,  Gavin & Gibbs, P.C.
                                                Washington, D.C.
                                                 Robert J.  Mason
                                             Chief,  Guidance Section
                                          RCLA Enforcement  Division
                                    U.S. Environmental Protection Agency
                                                Washington, D.C.
  I. INTRODUCTION  TO  LIABILITY  UNDER  CERC-
    LA/SARA
       • —
    A. Factors That Impaired Ability of Response Action Con-
       tractors (RACs) to Offset Liability Risk
       • Strict, joint and several liability under Superfund and
         some state laws
       • Lack of adequate and affordable liability insurance

    B. Pollution  Liability Insurance Industry  —  Lack of
       Coverage
       • Pollution  liability  insurance  has  been generally
         unavailable
       • Lack of insurance has discouraged qualified firms
         from engaging in Superfund cleanup

    C. Pre-SARA Indemnification of RACs
       • EPA Contract authority under CERCLA
       • Indemnification above an initial SI million for third-
         party liability

 II. POLICY REASONS FOR FEDERAL
    INDEMNIFICATION OF RACS

    A. Incentive to  get  RACs to work on Superfund projects

    B. Encourage insurance industry to become involved

III. CONTRACTOR LIABILITY UNDER SUPERFUND

    A. Negligence standard of liability under federal law
       • Liability under CERCLA or any other Federal law
         resulting from a "release or threatened release"  is
         limited to  damage caused by negligence, gross negli-
         gence or inten.tional misconduct
       • Contractor relieved from  "strict"  no-fault  tort
         liability under federal law

    B. No federal preemption of state and common law using
       different liability standards such as strict liability —
       beware of strict liability under state law

    C. Employees of State and Political Subdivisions who pro-
       vide services at Superfund site have the same liability
       exemption as provided to RACs under SARA

    D. Potentially  Responsible Parties (PRPs)  cannot use
       SARA Section  119(b) as defense to  their Superfund
       liability

520    SEMINARS
IV.  INDEMNIFICATION PROVISIONS

    A. Authority
       • SARA constitutes the sole authority to indemnify
         RACs and places specific limits on the authority
       • Executive Order 12580 delegates the President's
         indemnification authority to the U.S. EPA and other
         federal agencies
         —  U.S. EPA is authorized to exercise its discretion
            and entering into indemnification agreements with
            RACs
         —  The indemnification authority applies from the
            date of enactment of SARA (Oct. 17, 1986). As
            inter-preted by EPA, the Section 119 provisions
            limiting the circumstances under which indemni-
            fication may be granted must be strictly adhered
            to starting with October 17,  1986
         —  Other federal agencies such as, the Department of
            Energy may execute their own indemnification
            agreements with RACs and must use their own
            appropriations to pay claims on those indemnifi-
            cation agreements
       • Executing an indemnification agreement is  discre-
         tionary with the federal agency
       • RACs may  only be held  harmless for costs and
         damages (including the expenses of defense and settle-
         ment) arising out of the RACs' conduct constituting
         negligence. Gross negligence and intentional miscon-
         duct cannot be indemnified. Nor will the U.S. EPA
         indemnify a RAC for a strict liability arising out of
         state  or common law

    B. Applicability
       • Indemnification will only  apply where a response
         action is carried out under written agreement between
         a RAC and the U.S.  EPA or other federal agencies,
         states (including state political subdivisions) which
         have entered into contracts or cooperative agreements
         in  accordance with  a ง104(d)(l) and with PRPs
         carrying out an agreement under a'al22 (relating to
         settlements) or a ง106 (relating to abatement)

    C. Source  of Funding
       • Indemnification costs will be considered governmen-
         tal response costs under a ง104 of Superfund and
         maybe recovered from PRPs under a ง107

-------
       • The Superfund (Hazardous  Substance  Superfund
         established under the Internal Revenue Code of 1954,
         Chapter 98, Subchapter A) is specifically identified
         as the source for funds for indemnification when the
         U.S. EPA is providing the indemnification
       • Other federal agencies which execute indemnification
         agreements with RACs must use their own appro-
         priations to pay claims. The U.S. Army Corps of En-
         gineers, when managing a project under interagency
         agreement with the U.S.  EPA, will not  be deemed
         another  "federal agency" and the funds  will, there-
         fore, come from the U.S. EPA

   D.  Requirements for Executing Indemnification Agreement
       • Liability exceeds or is not covered by insurance avail-
         able at a fair and reasonable price, and adequate in-
         surance to cover such liability is not generally available
         at the time the response action contract is entered into
       • RAC must have made and must continue to make
         diligent efforts to  obtain insurance
       • Where RAC contract covers more than one facility,
         the RAC must agree to continue to make diligent
         efforts to obtain insurance each time the contractor
         begins work under the contract at a new facility

   E.  Limitations on Indemnification
       • Only liability resulting from a "release"  arising out
         of a response action activity may  be indemnified.
         Section 119(c)(5).  The scope of coverage therefore
         applies  only  to liability arising from sudden,  and
         accidental and  gradual pollution releases
       • Deductibles and limits will apply to all indemnification
         agreements
       • Indemnification agreements with RACs working for
         PRPs—  the U.S.  EPA  must first  determine what
         amount  of indemnification the PRP is  capable of
         providing to the contractor. EPA can only enter into
         an indemnification agreement if it determines that the
         PRPs are unable to provide adequate indemnification
         to cover "any reasonable potential liability of the con-
         tractor . ..." In making such a determination, the
         U.S. EPA must take into account "the total net assets
         and resources of potentially reasonable parties with
         respect to the facility at the time of such  determina-
         tions."  Section 119(c)(5)(C)(i).
       • Conditions on  indemnification working for RACs
         working for PRPs
         —  RAC must exhaust all administrative, judicial and
            common-law claims for indemnification against
            all PRPs participating in the cleanup of the facility
         —  Before recovering under  the  indemnification
            agreement,  the RAC must pay the  deductible
            which was established by the agreement

   F.  RCRA facilities  may not  be  indemnified,  a
       ง119(c)(5)(D).

   G.  Persons retained or hired by RACs may be  eligible for
       indemnification, provided the U.S. EPA specifically ap-
       proved of the retaining or hiring of such individuals.

   H.  Cost Recovery — Indemnification claims are considered
       cost of response actions recoverable from  PRPs.

V. EPA IMPLEMENTATION OF SARA PROVISIONS

   A.  General Requirements for Obtaining Indemnification.
       1. RACs will be required to submit written documenta-
         tion to  the U.S. EPA concerning  efforts made to
         obtain pollution liability coverage. The RAC will also
         have to periodically submit written documentation
          demonstrating additional efforts it has made to secure
          insurance.

        2. A RAC that obtains liability coverage must submit
          a copy of policy and declaration page to the U.S. EPA
          or the contracting officer.

     B.  Indemnification of RACs working for the U.S. EPA —
        the U.S. EPA will enter into new indemnification agree-
        ments with RACs who are currently working under con-
        tract with the U.S. EPA for work which they will initiate
        at a new  site after the date of enactment of SARA.

     C.  Indemnification of RACs working for states — the U.S.
        EPA will not offer indemnification for RACs for site
        work  they performed for states  prior to the date of
        enactment of SARA. It is likely that indemnification will
        only be offered if the response action is part of new work
        initiated at the Superfund site after SARA enactment
        and is directly related to the  cleanup of the site.

     D.  Indemnification of RACs working for other federal
        agencies—Federal agencies other than the U.S.  EPA
        which have been delegated authority to indemnify con-
        tractors will exercise their own discretion, and the funds
        for  indemnification  will be  drawn  from their own
        budgets.

     E.  Indemnification of RACs working  for PRPs
        • The U.S. EPA will only indemnify RACs performing
          response action activities for PRPs subject to a con-
          sent order or decree at Superfund sites after date of
          enactment of SARA.
        • The U.S. EPA will only use its authority to indem-
          nify RACs working for PRPs if adequate indemnifi-
          cation cannot be provided by the PRPs and as a result
          PRPs are unable to obtain  the services of qualified
          RACs.

     F.  Publicly owned treatment works subject to permit-by-
        rule provisions cannot be indemnified by the U.S. EPA.
        ง119(c)(5)(D). This provision is intended to prevent the
        U.S. EPA from offering indemnification to off-site
        treaters or disposers of Superfund  Hazardous Waste.

 VI.  STATE LIABILITY STANDARDS  FOR RACS—
     A report  on  the statutory, common law and contractual
     liability standards at state level.

VII.  RISK MANAGEMENT TO REDUCE
     POTENTIAL LIABILITY

     A. Risk Management Practices
        • Identify hazardous substances that may potentially
          be released
        • Assess potential areas of liability associated with site
          activities
        • Document on-site conditions and operations
        • Compare site conditions via an independent verifi-
          cation of compliance with statutory, regulatory and
          operational requirements
        • Assess  corporate procedures, policies and guidelines
          related to hazardous waste management
        • Examine environmental risk management practices
          for  accuracy  of  reporting  and recordkeeping
          programs.

     B.  Quality Control
        • On-site health monitoring
        • Incident management  planning and techniques
        • Establish a plan to deal with uncertainties
                                                                                                        SEMINARS     521

-------
      C.  Considerations in Selecting Project Work
          • Does the cleanup involve a NPL site?
          • Is it an orphan site?
          • Is offsite disposal required?
          • How close are populated areas?
          • Is  contractor  technically  capable  and  is  client
            financially stable?

      D.  Considerations in Selecting Off-Site Cleanup Options
          • Reduce waste  volume — precondition the wastes
            on-site
          • Transportation concerns and problems
          • Selection of location for  disposing wastes

 VIII. OTHER WAYS TO  POTENTIALLY REDUCE
      LIABILITY RISKS

      A.  Forms of Business Organization
          • Incorporation  of  Sole  Proprietorships  and
            Partnerships—Design professionals such as architect
            engineers may not be able to avoid personal liability
            by incorporating. Other response action contractors
            may receive some degree of protection by incorpora-
            tion. State statutes and common law doctrine should
            be carefully reviewed before making a decision  to
            incorporate.
            —  Formation   of subsidiaries  without  proper
               guidance and rigid adherence to the practices dic-
               tated by the law of the applicable jurisdictions,
               there is potential that a subsidiary  will be deemed
               a "mere instrumentality" of the parent corpora-
             tion and that iis veil will be pierced. Attention
             should be given to assuring sufficient capitaliza-
             tion; separate financial records/reports and insti-
             tutions; separate board and shareholder meetings;
             separate staffs; formal arms-length agreements
             regarding the use of assets and purchase services.

    B.  New Contract Language Used to Reduce or Shift Risks
        •  Indemnification Clauses. A clear and unambiguous
          indemnification provision in a contract between two
          commercial parties of  equal bargaining power may
          be  upheld by the courts absent  a statute or public
          policy prohibiting such an agreement. The majority
          of  states, however, have enacted  anti-indemnity
          statutes.  Even if  an indemnification provision is
          enforceable, it  is only worth what the indemnitor is
          worth. RACs may, therefore, need to investigate the
          financial capability of the party with whom they are
          contracting.
        •  Limitation of Liability Clauses.  A limitation of lia-
          bility clause establishes a limit to the liability that will
          be  incurred. Such a clause may survive judicial
          scrutiny where an indemnification clause would not.

IX. CONCLUSION

    A.  Partial  Solutions to the  Liability  Insurance Problem
        Provided by Congress

    B.  Solutions Within the Control of Potential Pollution Lia-
        bility Insurance Industry

    C.  What This Means to the  Response Action Contractor
522    SEMINARS

-------
         Soil-Gas  Surveying  for  Subsurface Organic  Contamination:
                                  Active and  Passive Techniques

                                                    H. B. Kerfoot
                       Lockheed Engineering  and Management Services Company, Inc.
                                               Special Projects Office
                                                  Las Vegas, Nevada

                                                P. B.  Durgin, Ph.D.
                                      U.S. Environmental Protection Agency
                                 Environmental Monitoring Systems Laboratory
                                                 Las Vegas, Nevada
INTRODUCTION
  Soil-gas surveying to delineate subsurface contamination by vola-
tile organic compounds (VOCs) is a technology that is finding wide
applicability in preliminary site-characterization and leak-detection
efforts. Because of the low cost of these techniques in comparison
to drilling wells or boreholes and sending soil or groundwater
samples to a remote laboratory for analysis, soil-gas surveying is
useful to  indicate sources of contamination and to plan the loca-
tions of monitoring  wells. However, since the technique is an
indirect one (i.e., measuring soil-gas concentrations  and not
groundwater concentrations), the analysis must always be con-
firmed by direct  measurements  of contaminants in the soil or
groundwater. In addition, because the technology depends upon
transport of volatile compounds through the vadose zone, vadose
zone conditions must be amenable to the techniques of soil-gas
surveying.

POROSITY
  For soil-gas surveying to be successful, a continuous vadose-/one
air-filled  porosity above approximately 5% is required to allow
sufficient transport of VOCs from the source through the subsur-
face. Because VOC transport through the vadose zone is by diffu-
sion, a linear increase in concentration with depth is anticipated
given steady state and homogeneous soil conditions. The steady-
state linear depth dependence of vadose zone gas concentrations
allows evaluation of depth profiles of concentrations to differen-
tiate between surface and subsurface sources of VOCs. Vadose-
zone layers with low air-filled porosity are barriers to diffusion
and can create very low concentrations above them and maintain
concentrations at high levels  beneath them. Such barriers  can be
subsurface (e.g., clay) or surface barriers (e.g., pavement). Infiltra-
tion of rainfall can fill vadose-zone pores and displace dissolved
gases, thus eliminating them from the soil atmosphere temporarily.

FATE OF CONTAMINANTS
  In planning, interpreting or managing soil-gas surveys, the fate
of VOCs in the vadose zone must be considered. Hydrocarbons
are prone to oxidation in the vadose zone and saturated zone. The
absence of these compounds has been observed in soil gases over-
lying contamination  at shallow depths where the rate of oxygen
supply from the atmosphere  is rapid. Most chlorinated solvents,
on the other hand, are not prone to oxidation but can undergo
reductive dehalogenization. This process typically occurs at a much
slower rate than hydrocarbon oxidation so that it does not usually
eradicate the presence of the target chlorinated solvent. In addi-
tion, the  products of reduction of chlorinated solvents are still
highly characteristic  soil-gas indicators of contamination.
VAPOR PRESSURE
  The compounds that are good targets for soil-gas surveying have
three requisite characteristics. First, a vapor pressure above 0.1
kPa is necessary. Second, a Henry's Law constant (vapor pres-
sure divided by water solubility) above 0.01 kPa.l/mg is required.
Finally, if direct detection of the compound(s) is proposed, the
tendency towards subsurface transformation must be evaluated.
For chlorinated solvents, the possibility of decomposition to other
chlorinated products should be evaluated in planning the survey
and evaluating results. For hydrocarbons (e.g., gasoline, jet fuel)
the possibility that oxidation may completely remove them from
soil gases should be considered. Under such circumstances, success
has been reported in measurement of soil-gas carbon dioxide (the
product of hydrocarbon oxidation) to delineate hydrocarbon con-
tamination. This technique also is applicable to fuels of lower vapor
pressure (e.g., diesel, JP 8,  bunker  fuel).

ACTIVE AND PASSIVE SAMPLING
  Soil gases can be sampled using one of two general approaches:
active sampling or passive sampling. Active sampling techniques
depend upon drawing soil gas through a sampling probe and
analyzing subsamples. This analysis typically is performed on-site
and provides real-time results. On-site analyses can be performed
using sophisticated gas chromatographs (GCs), portable GCs or
other techniques depending on the data quality objectives of the
survey.
  Passive sampling of VOCs from soil gases relies upon diffusion
of the VOC from the soil gases to a sorbent  surface. Although
passive sampling does not give real-time results, the technique does
not require sophisticated equipment or analytical expertise on-site,
and the cost is typically lower than that of active sampling methods.
  When the sampling strategy will evolve on the basis of real-time
results, active sampling with on-site analysis is the technique of
choice. However, when the goal of the survey is to clarify the situ-
ation between wells or other cases where a fixed-grid sampling net-
work will be  used,  passive sampling can be a cost-effective
alternative that can yield more precise results.
  The choice of the approach used  in a soil-gas  survey depends
on site characteristics, level of knowledge of the situation and target
compounds. A survey with no knowledge of existing contamina-
tion should be performed with the sampling and  analytical flexi-
bility of active sampling with on-site analysis. In situations where
some knowledge of the contaminant identifies is available, passive
sampling with remote analysis is a viable option. Aquifers with
hydrocarbon contamination are good candidates for indirect
detection of contamination through soil-gas carbon dioxide meas-
                                                                                                     SEMINARS     523

-------
urement. In cases where oxidation of hydrocarbons may occur,
deep soil-gas sampling  (below 10 ft) can  avoid the  problem.
Sampling locations should be selected  to  address  the specific
reason(s) for conducting the survey. When high concentrations or
a pavement-covered vadose zone are encountered, samples should
be further apart. At lower dissolved VOC concentrations in rela-
tively porous soils, sample locations can be  spaced more closely,
at separations down to the depth of the VOC source.

QA/QC
  QA/QC for soil-gas surveys can include several techniques. In
order to demonstrate that soil-gas measurements reflect subsur-
face conditions at the site, results should be compared to tradi-
tional groundwater or soil-core analyses. Closely spaced samples
can be analyzed at one or more location(s) to provide an estimate
of the overall precision of results so that results can be  evaluated
with an idea of what constitutes a meaningful difference in results.
Depth profiles can  be  included in the study for differentation
between surface and subsurface VOC sources.
DATA ANALYSIS
  Data interpretation from soil-gas surveys often  is performed
using contour maps. Linear interpolation, geostatistical and Fourier
transform techniques have been used successfully. Before contour
maps are prepared, the data should be evaluated for each VOC
to ensure  that such an effort is worthwhile. When both  depth-
profile and area! information are available, the vertical and horizon-
tal gradients can be used to generally indicate sources. However,
as with any screening technique, care should be taken to avoid over-
interpretation of results. Subsurface and surface barriers to diffu-
sion such as clay lenses, perched waterbodies, buildings and paved
areas as well as episodic factors such as water-table fluctuations,
rain or surface-soil freezing can alter soil-gas survey results.

ACKNOWLEDGEMENT
  The information in this document has been funded in pan by
the United States Environmental Protection Agency under contract
number 68-03-3249 to LEMSCo. It has been subjected to Agency
review and approved for  publication.
524    SEMINARS

-------
                       Introduction to Dispersion Modeling
                                 Of Hazardous Releases
                                      Seminar  Outline
                                   Ashok Kumar, Ph.D., P.Eng.
                                  Department of Civil Engineering
                                     The University of Toledo
                                          Toledo, Ohio
Topic
                 Course Outline
      INTRODUCTION

      1.1 Hazardous Releases
      1.2 Regulations

      AIR QUALITY MODELING

      2.1 Basic Meteorological Processes
      2.2 Source Term
      2.3 Dispersion Models: Continuous
        &Puff
      2.4 Handling of Heavy Gas Releases
      2.5 Dispersion Coefficients

      INPUT DATA REQUIREMENTS

      EXAMPLES

      USE OF PERSONAL COMPUTERS
                                                                        Table 2
                                                               Evacuation Size and Distribution
  Time

10 Minutes




90 Minutes
10 Minutes

10 Minutes

10 Minutes
                                        40   60    SO
                                          No. of lncia.nl.
             Table 3
   Progress of U.S. Federal Law Regulating
   Hazardous Material Spills as Published
In the Federal Register by U.S. EPA (1978-1980)
Table 1
Chemical Accident Evacuations by Cause
Cause of evacuation
Train derailment
Train car spill/fire
Truck accident
Truck spill/fire
Chemical plant release
Industrial plant release
Pipeline
Ship incident
Waste site accident
Other
Totali
1980
14
3
9
1
5
3
2
2
0
4
43
1981
8
6
9
11
10
10
1
1
1
5
62
1982
13
5
6
4
15
18
1
0
2
4
68
1983
12
4
6
9
8
23
0
0
3
0
65
and Year
1984
8
5
5
7
5
24
0
1
1
1
57
Total!
55
23
35
32
43
78
4
4
7
14
295
Data Rules
Dec. 11, 1973 Oil spill prevention
Aug. 22, 1974 Designation and determina-
tion of removability
harmful quantities, penalty
- rate
Mar. 13, 1978 Hazardous substances
Sept. 1, 1978 SPCC Plan; best management
plan
Aug. 29, 1979 Reportable quantities
March 19, 1980 National contingency plan
update
•Number of hazardous chemicals on the list.
Chemical
Oil
Haz. mats.
(370)'
Haz. mats.
(306)'
Haz. mats.
(299)ซ
Haz. mats.
Haz. mats
(299 )•
Oil & Haz.
mats.

Action
Spill prevention
Proposed rules
Proposed rules
Proposed rules
Proposed rules
Final rules
Most recent
revision

                                                                                   SEMINARS   525

-------
            i SPILL SITE
WIND.        |  -••-..
DIRECTION    I        "••••...                 W/2
(W or 270')    |	_":_•_• •^-••_. .L.^---'	1
            I	—	K —
                            Figure 1
           Hazard Zone Mapping For Wind From 270ฐ C
   WIND
   DIRECTION    '
(from W^r         — —
     270' - Id')
                                                             X (km)
                            Figure 4
       JANUARY AND Jl'LV  WIND ROSES, CINCINNATI
The monthly distribution of wind direction and wind speed are summarized
on polar diagrams. The position] of the spokes show the direction from
which the wind was blowing; (he length of the segments indicate the per-
centage of the speeds in various  groups.
                            Figure 2
           Hazard Zone Mapping For Wind From 270 ฐC
                                     I[C corrltfei t*ปlai


                                     lnd  .•rlซDUItT  lit
ouu

500

400
HEIGHT
ABOVE
GROUNO(m)
300
200
100

0
1 l i I I l l l | i :
GRADIENT WIND f :
' /"
DAY/ /
/ •'' ~
/
I/ '
.--/
NIGHT ..-••''' S
	 ' ' "SURFACE" WIND
-i-:JJ_.-i — r — I l l i 1 l i
31 23456789IOI
                             Figure 3
     Toxic Corridor Forecast Worksheet With Sample Calculations
                                                                                                      WIND SPEED (m/sec)
                             Figure 5
     CHANGE OF WIND SPEED PROFILE WITH STABILITY
 The frictional drag reduces the wind speed close to the ground below that
 found at the gradient level. The profile at  night when the air is stable is
 usually steeper than that found during the day.
526    SEMINARS

-------
                                   SUPERAOIABATIC
                                           DRY AOIABATIC
                                             LAPSE RATE
20 21 22


IOO




O
V.
s
\
N
N
"

i


ISOTHERMAL
•* — '

S
\
V , |
                                                                        Table 4
                                        Some factors influencing spilled chemical input into the environment
                                                                                                                     Evaporation Rate
                                                                                         Solar    Heat
                                                                                         heating  transfer to
                                                                                                surroundings
                                                          Chemical   Spreading  Evaporative  Wind  Temp   Supply of
                                                          properties  on        cooling                   chemical
                                                                    surface                           (leak rate)
                                                                                                                  Penetration
                                                                                                                  Surface characteristics
                                                                                               Chemical properties
                                                                                               Tank configuration
                                                                                               Tank position
                                                                                               Tank geometry
                                                                                               Insulation
                                                                                               Gravity
                                                                                               Sonic limits
                                                                                               Temperature and wind
                                                                                               Heat transfer
                                                                                               Evaporation rate
                                             21
                              19        20
                                   TEMPERATURE CO
                                     Figure 6
               TYPICAL ENVIRONMENTAL LAPSE RATES
    Typical examples of vertical temperature profiles are shown in compari-
    son with the dry adiabatic lapse rate (— 1C/100 m) which  serves as a
    reference for distinguishing unstable from stable cases. The position of the
    dashed line representing the adiabatic lapse rate is not important; it is sig-
    nificant  only as far as its slope is concerned.
Instantaneous point  sources
Continuous/semi-continuous point/
        area sources
                                                                       Table 5
                                                  Estimates of error if certain factors are ignored
Factor
Leak rate models
Tank position,
geometry
Tank insulation
Heat transfer
Evaporative
cooling
Molecule speed
limits
Rupture size
Evaporation models
Evaporative
cooling
Wind
Spreading on
surface
Heat transfer to
surroundings
Solar heating
Estimate of
average error]

1 to 10
1 to 10
1 to 10
1 to 3
1 to 3
1 to 2

1 to 2
1 to 2
1 to ID2
1 to 1.S
1 to 1.5
Estimate of
maximum error]

10?
10*
10*
101
102
5

103
10
10*
3
3
Direction
of erroT2

Less
Less
More
Less
Less
Less

Less
More
Less
More
More
Examples of
"worst" cases

Horizontal
tank in vertical
position
Tank
completely
insulated
In fires
Volatile
chemicals
Apparent
rapid leaks
Highly Irregular
openings

Volatile
chemicals
High wind
situations
Contained
spill
Cryogen on water
Spill on sunny
day
                                     Figure 7
                       Decision Tree For Model Selection
                                        1.  Errors are given as factors differing from the present value (e.g., 1 is no error, 2 Is esti-
                                           mated to be as much as twice, or half, the actual value, etc.).

                                        2.  Less Implies less chemical Input to the environment than simple models predict.
                                                                                                                                          SEMINARS     527

-------
    S
IU
5
2
103
5
2
W
5
2
101

10ฐ 1














/
/ ,
'//
'//
' //
cV-
^ —












*
/
/
' /
/
y
/















t
s

t
/
/















S

s
/
/













/ /
2

^ /
^ t
ฃ












/
//
'•'//
:••/-/
' S
' /
, '











j
/
/
/
',
/ -
K
B






S
-?zz

^2
2 i
//
' ,'
/

- EXTR
- MODE
C - SIW
D - NEUT
E r SUGf
F - MODE









V,
//•/
!' C_
! . - -7
:;#!
< * '
,''s*~
( I '






/
/,
^
—^
^^





;MELY UNST
RATEIY UNS
UY UNSTAB


^ ^
2 4
^/;
/ ^
//x
/ /







iRIF
fABLE
LE
RAL
WY STABLE
RATELY STABLE.






. p
. , !


. ^
, '
'













            10s   2      5    103   2      5     10*   2
                            DISTANCE  FROM SOURCE (m)
10s
                                           Figure  10
                                 Schematic Contaminant Plume
                                Figure 8
           Lateral Diffusion, ay.  Versus Downwind Distance
             From Source for Pasquill's Turbulence Types
                              (Slade  1968)
3 • 103
2
103
5
J
ฃ 2
TICAL DISPERSION CO
5.. ~
CC
ป 5
b"
2
10ฐ











/*


^/
r S
/









f
/ (
' /
/

f








\
f

' 7
* /
7 /
^ ^
"^ri i





/

"T —

' >
/ /
-tc^
C

, ^\x^
'..',,

	

••'
'
'



-/
t


/

C ^

x-^


4





S

J

TJ
f
/


j
/
,/
"'
^
r*w

TRFW
3 - MODERi
: - sudHii
) - N[UTRA
1 - SLIGHTL
- MODERA





-'•'- x<
f


'X
^^-*
• •"


ELY UNS'


^
^
***



ABI
fTETY UNSTA
Y UNSTABLE
L
Y STABLE
fELY STAI
III

LE

X1



<
-




Y
BL





^'


^

>•
— t-j-




•'. 11




10* 2 5 103 2 5 10* 2 5
DISTANCE FROM SOURCE (m)




„ i


T '










105
                               Figure 9
           Vertical Diffusion, oz.  Versus  Downwind Distance
             From Source For Pasquill's  Turbulence Types
                              (Slade 1968)
                                                                                      DETERMINE TOTAL AMOUNT
                                                                                            DISCHARGED
                                               SUp I. Ute Figure

                                               Time since rupture	
                                               Equivalent diameter et rupture —__ mm
                                               Percent el chemical remaining	%
                                               Amount discharged:
                                               q = ซOOOOL-%"ปOOOOL =	- L
                                              __ L • density ptf/L) t 10OO - 	 tonnes
                                                                                      CALCULATE POOL RADIUS (r)
                                               Sltp 2: 2 mm peel ihicknett

                                                         l - - _. km
                                                                                 CALCULATE VAPOUR EMISSION RATE (Q)

                                                                                                i
                                                                                     DETERMINE WIND SPEED IU)
                                                                                         AND DIRECTION |D|
                                               Step 3: Uiซ Figure
                                                        0	f/i

                                               Sup 4: Observed er eitlmaled
                                                        U =	km/*; D = .
                                                                                    DETERMINE WEATHER CONDITION
                                           J   Step S.
                                                                                                                        Use Table    Condiliปn =
                                                                                  DETERMINE HAZARD CONCENTRATION
                                                                                    (C) • LOWER Of LFL ซ TLV"ซ 10
                                                    : C = 0 05 g/ซ* lor Nitric i
                                                        10 " TLV^1983)
                                                                                      [   COMPUTE C • U i Qj
                                                                                                                  Step ? C--
                                                                                                                            CU/Q '
                                                                                   CALCULATE HAZARD DISTANCE Ha I   Sup *: Ut* fl*"r€
                                                                                    FROM VIRTUAL POINT SOURCE   |            t -     fcn,
                                                                                   CALCULATE HAZARD DISTANCE X.
                                                                                   FROM AREA SOURCE X, = Xp-10,
                                                                                                                  Step 9: Compulation required
                                                                                     CALCULATE PLUME HAZARD
                                                                                       HALF -WIDTH (W/21 m>,
                                               Step 10: Ute Table

                                                         <*"*> ma.. -
                                                                                    DETERMINE TIME III SINCE SPILL |    Step 11    t = .
                                                                                  CALCULATE DISTANCE (Xt) TRAVELLED I  *'•ป 12; u" Fiซu"     wlln U '""" Sttป **
                                                                                 BY PLUME SINCE TIME (II OF ACCIDENT |            i . 	km
                                                                                     HAZARD ZONE AND PLUME
                                                                                        LOCATION DEFINED
                                           Figure 11
                        Flow Chart to Determine Vapour Hazard Zone
528     SEMINARS

-------
                            Time from Puncture (mm)
                                 Figure 12
                       Percent Remaining vs. Time
   1000
                           Time Irom Puncture (mm)
                                 Figure 13
                         Discharge Rate vs. Time
         10"











.."..:
























29 "
























10
























^8
























/s




	 ฃ.


v/.

!~~1~
:
: ;
/











	


/
I
'//
ti
,
(
f
— y
-/
/














V/
/ /
/ '
//
//
•/-
j
/
/












\\-A
~~h
//
'I
'//
/V-v
'Ft
U
l\l
••/
!'•
;
•
i
;
:








f
* r
f



j
.*
.•













•'



!/
|/


















- 20'C
' 10'C

' O'C

1 -10'C




- *ซol
lor
See
lor
com
M
Ha
tke
•10





                                                      * Note: Nomogram applies
                                                        lor wind speed ol 4 5 *>A
                                                        See Introduction Manu.it
                                                        lor relationships  to
                                                        compute vihies lor othr
                                                        m'mi  speeis. it necessary
                                                        *so.  He w&I portion; DT
                                                        tie cunts represent ;i'
                                                        •IOCS to 120 tonnes
                                     100 >        1 O'JO
                                        r = 120 m
                         Liquid Pool Radius . r (m)


                                 Figure 14
Vapour Emission Rate vs. Liquid Pool Radius For Various Temperatures
                                                                                               10
                                                                                                   0.1
                                                                                                                            X  = 8.5 km1
                                                                                                                   Downwind Distance, X (km)
                                                                                                                      Figure 15
                                                                                            Normalized Vapour Concentration vs. Downwind Distance
                              Table 6
              Maximum Plume Hazard Half-Widths
                (For Nitric Acid (42ฐBe)  At 20 ฐC)
Vcalher Condition D
0/U (*/2)maป
(B/m) (m)
3>0 000
300 000
ZJOOOO
200 000
175000
150000
125000
100000
75000
50000
JO 000
25000
20000
15000
10000
7 500
5000
2 500
1 000
750
500
250
100
50
3430 (99.5km)'
3115
27J5
2425
22J5
2030
1315
15SO
1320 Q/U= I3SO-
1030
750
ฃ70
Hi
500
395
335
260
175
100
85
70
50
30
20
Weather Condition F
P'/U, i*{2W
(g/m) (m)
30000 1430 (99.5 km)'
25000 1250
20 000 1060
1 5 000 S50
10 000 630
7 500 510
5 000 375
2 500 200
1 500 175 - W/2mlx = 175 m
1 000 135
750 110
500 S5
250 55
100 35
X 2)


• Data are provided up to a
maximum downwind hazard
distance of 100 km.




I'.Kjmplc! A spill releasing nitric acid vapour at the rate of Q = 2.9 x 103 g/s under
        weather condition F and a wind speed U = 2.1 m/s means Q/U s 1380 g/m, which
        results in a maximum plume hazard hall-width (W/Zlmuc = I7J m.
        Above table is valid only (or a nitric acid concentration ot lOxTLV*  or
        0.05 R/mJ.
                                                                                                                                        SEMINARS      529

-------
       Wind Speed (U)
 liquid Spill Site  (g)
                     Inilnl Vipour Pull
                                         Downwind Distance (I)
                              Figure 16
                   Schematic of Contaminant Puff
                                                                                                     Tien lnซ f MClwf (•")
                                                                                                           Figure 18
                                                                                                   Percent Remaining vs Time
                                                                               tow
                                                                                                     Tot torn r*actsปt
          Figure 19
   Discharge Rate vs. Time



ACCIDENT
LIQUID SPILLED



DETERMINE TOTAL AMOUNT
DISCHARGED


r ,

r

DETERMINE WIND SPEED (U)
AND DIRECTION (D|
^
DETERMINE WEATHER CONDITION
|
DETERMINE HAZARD CONCENTRATK
(O • LOWER OF LFL or TLVปป 10
i
[

COMPUTE CซQ,
\
r


100
Sttp 1 Ust Figure 90
Equivalent diameter of puncture ™— , mm ;Q
Percent of chemical remaining 	 \ ,-
Amount discharged. ** "
q = SO. 000 L * ซ 60.000 L = L T *ฐ
	 tonnes * 10* crams/tonne = 	 era mi * in
- 	 T "
Step 2: Observed or estimated ฃ M
0
Step 3 Use Table 0
Condition = 	

Hydrogen Chloride
10 > TLV* (19111)
Slop S: Computation required
C/Q, - 	 m '
CALCULATE HAZARD DISTANCE FROM I ป"P 0; u" rigute
INSTANTANEOUS POINT SOURCE I x ซ km

J
r

CALCULATE HAZARD
HALF-WIDTH (W/2|m,x.
|
DETERMINE TIME |t) SINCE SPILL
CALCULATE DISTANCE (Xt) TRAVELL
BY PUFF SINCE TIME (t) OF ACCIOEF
Step 7; Use Table
(W/2) .. = 	 m

— — —
V
Xs
i5\'
\*
i



5=
S,
S
\
s^
s\
\
\

=
— .
—
^
K
s

\
V

\
^ s
N""r'5
X ^
i^ SL S
\ '\U
vJV.lJ
v$K%N
\V^\ m^
	 	 —
* , ^"x,

-i? Nป
M^\
t'l i!
WL . t 	
JVC!
If. i \
Eli lA
i
& i
>


\
s


>
w




\

f


\






^



\





s N
* ,



^
^
^
J
1'!

i-
',






i ,'0 wo
\mt tl Fsactm (em)
Figure 20
Percent Remaining vs. Time
Pปซclปrซ M lollm ef In* Cat



Slop 6; r .uuo -
M
-1 ' - 	 • 2
* 3so -r
10 Step 9: Use Figure ฐ
IT with U from Steo 2 ป —
1 X, = 	 km =
LOCATION DEFINED
l_







!
















/





	 ^z.



*


^







*f







,-"J














— -







-^- ==*






Ls-^







- _ 	 M |M )JO ^ ?M J(|0 JM j(m ^ ^
                              Figure 17
            Flow Chart to Determine Vapour Hazard Zone
           Figure 21
Discharge Rate vs. Puncture Size
530     SEMINARS

-------
                          Puncture in Top ol Tantcar
                                                           100
                         Time from Puncture (min)
                                Figure 22
                      Percent Remaining vs. Time

                          Puncture irt Top of Tankcar
        50     100   ISO    200    250    300    350    400"  450    500
                    Equivalent Diameter of Puncture (mm)

                              Figure 23
                  Discharge Rate  vs.  Puncture Size
                               Table 7
               Maximum  Puff Hazards Half-Widths
                       (for hydrogen chloride)
Weather Condition 0
 QT
 (tonnes)

 2 230
 2000
 1 500
 1 000
   750
   500
   400
   300
   200
   100
   75
   50
   25
   20
   10
    7.5
    5
    2
    I
    0.73
    0.50
    0.20
    O.I
    0.075
    0.05
    0.01
Example;
          W2W
          (m)
                              Weather Conditio
                                        (W/2
                                        (m)
Or
(tonnes)
   4030  (99.4 km)ซ             100        1825 (97.1 km)'
   3 850                        75        1 610
   3 450                        50        I 360
   2950                        25        I 010
   2650         Qr=20 tonnesป    20         915  ซ(W/2)max = 915 i
   2 265                        10         680
   2080                         7.5       600
   I 860                         5         505
   I 600                         2         350
   I 223                         1         270
   1 010                         0.75      240
    940                         0.50      205
    720                         0.20      145
    670                         0.10      110
    520                         0.075     100
    4<5                         0.05       85
    400                         0.01       15
    285
    220
    200
    170
    120
    95
    85
    75
    40
Under weather condition F and QT = 20 tonnes, the puff hazard half-width
WZlmax = "3 m
Above table is valid only for a hydrogen chloride concentration ol 10 x TLV*,
or 0.07 g/m'.
                                      * Data are provided up to a maximum
                                      downwind hazard distance of 100 km
                                           O-
                                           CJ
1U


io-3


io-4


io-5


10'6


io-7


io-8
3.5 ซ 10 9—
io-9


io-10


in-11

^
\
\
\-
\
>
























V
^
i
3 :
\
!














	 T











\
m

\
fr


i —






'


i ....












V

L
\
- V

























.
l[[[[\ I I
\
' \
1 hr
- -L







_.. i

















_ .
-, .
~-

^




                                                                                                             F CLASS
                                                                                                                                                   D CLASS
                                                                                                         0.1
                                                                                 .0
                                          10
                                                                                                                                               100
                                               X = 42  m
           Maximum Downwind  Hazard Distance,  X (km)
                          Figure  24
Normalized Vapour Concentration vs. Downwind Distance
                                                                                         -5
                                                                                                             5         10        15
                                                                                                              TEMPERATURE  (ฐC)
                                                                                                                  Figure 25
                                                                                                          Phase Diagram of Water
                                                                                                   20
                                                         25
                                                                                                                                      SEMINARS     531

-------
              Clinical  Medical Surveillance  and Epidemiology:
                   Methods for  Assessing and Protecting the
                       Health  of Hazardous  Waste Workers

                                                 By
                                Bertram W. Carnow, M.D., F.C.C.P.
                                Shirley A. Conibear, M.D., M.P.H.
COMPARISON OF MEDICAL SURVEILLANCE
AND EPIDEMIOLOGY

Medical Surveillance
1. Definition
2. Uses of Medical Surveillance
  a. Establishing a baseline
  b. Detecting Pre-existing Disease
  c. Determining Fitness for Employment
  d. Complying with Regulations
  e. Detecting Work Related Illness
  f. Generating Data for Epidemiologic Studies

Epidemiology
1. Definition of Epidemiology
2. Types of Studies
3. Uses of Epidemiology
4. Practicality of Carrying Out Studies
5. Cost
HOW TO INTEGRATK EPIDEMIOLOGIC TECHNIQUES
AND METHODS INTO A MEDICAL SURVEILLANCE
PROGRAM FOR HAZARDOUS WASTE WORKERS

Study Designs
1. Protocol Setting
2. Test Selection
3. Testing Sequences
4. Population to be Tested

Data Gathering
1. Medical Testing and Examinations
2. Standardization of Multiple Site Testing
3. Record Handling
  a. Design of forms
  b. Computerization of Data
  c. Quality Control

Analysis of Data
1. Controls
  a. Personal
  b. Group
2. Statistical Methods
3. Interpretation of Data

Uses of Epidemiologic Findings in
Defining Medical  Surveillance Programs
532   SEMINARS

-------
                      Bioremediation of Hazardous  Waste Sites:
                    The  Use of Biological  Treatment Techniques
                     For the  Detoxification  of  Sludges and  Soils
                                  Contaminated  with  Organics

                                            Raymond Loehr, PhD.
                                      The University of Texas at Austin
                                                 Austin,  Texas
                                                   John Ryan
                                   Remediation Technologies, Inc.  (ReTeC)
                                               Kent, Washington
                                               Ronald Linkenheil
                                   Remediation Technologies, Inc.  (ReTeC)
                                            Fort Collins,  Colorado
ABSTRACT
  This seminar reviews the current state of the art of bioremedia-
tion (biological treatment) of soils as a means to reduce the con-
centration of, and immobilize, organic constituents of hazardous
wastes and spill residues. The seminar:
• reviews the principals and theory of biological soils treatment
• indicates existing utilization of this technology and related bio-
  logical treatment techniques
• discusses regulatory and institutional constraints on, and incen-
  tives for  use of, the technology
• identifies the technical aspects of the design and operation of
  bioremediation facilities
• presents several case studies of bioremediation at Superfund sites
  and RCRA facilities
  The seminar summarizes the actual and potential uses of the tech-
nology, presents a  decision matrix for  the evaluation  of the
applicability of bioremediation at a site and makes recommen-
dations regarding the institutional and regulatory constraints which
could limit  its implementation.

A. Theory & Principles  of Biological Treatment

    1. Principals related to microbial degradation of organics in
      contaminated soils

    2. Contamination results from spills, leaks from impound-
      ments and other management practices related to storage
      and transfer  of organic chemicals and fuels

    3. Organic chemicals which  are candidates  for biological
      treatment include:
        volatile organic carbon compounds (VOCs)
        polynuclear aromatic hydrocarbons (pAHs)
        halogenated organics

    4. Types of sites which may contain these constituents
        petrochemical
        coal tar chemical
        petroleum  production, refining and marketing
        wood preserving
        electric and gas utilities

    5. Bioremediation in saturated vs. unsaturated soils

    6. Aerobic  vs. anaerobic bioremediation

    7. Factors effecting bioremediation
        soil moisture
     aeration and gas transfer
     loading rates
     toxicity and other characteristics of contaminants
     nutrients (or lack thereof) in soils
     amendments for
       organic material
       pH control
       microorganisms
       bulking agents

 8. Treatability studies to determine the operating conditions
   needed for successful performance of biological treatment
     design of the study
     bench scale  vs.  field
     variables
     parameters to be measured
     toxicity testing
     duration and cost

 9. Degradation and loss rates for selected chemicals
     half lives and reaction rates calculations
     treatability time and degree of remediation

10. The role of sorption and volatilization in bioremediation
Existing practices and Applications of Biological Treatment
for Sludges and Soils
Review of existing applications of biological treatment
 1. Land treatment
     Petroleum refinery wastes...solid & hazardous
     Superfund sites & RCRA closures
       manufactured gas plants
       wood preserving plants
       coal tar chemical plants
       agricultural wastes
     Existing facilities
       design & construction features
       operating features
       future regulation
     Performance and economics
     Future of the technology

 2. Liquid/solids  systems
     Wood preserving wastes
     Existing applications
       design &  constriction features
       operations
       future regulation
                                                                                                 SEMINARS    533

-------
c.
D.
      Performance & economics
      Future of technology

 3.  Waste water and contaminated ground water treatment
      Applications
         Municipal & industrial waste waters
         Contaminated ground and surface water
         Various subsets of the technologies
           aerated & facultative lagoons
           activated sludge
           trickling filters & other contact reactors
         Groundwater treatment
           extraction
           treatment
           reinjection
         In situ soils treatment
      Performance & economics
      Future of the technology
 4.  Composting
      Applications
        sewage treatment sludges
        agricultural wastes
        petroleum contaminated soils
      Facility features
      Operating techniques
      Performance and economics
      Future of the technology

 5.  Soil bio filter
      Applications
        air and gas scrubbers
      facility features
      Operating techniques
      Performance and economics
      Future of the technology
Regulatory Overview and Constraints

 1.  Summary of Superfund amendments that relate to the use
    of biological treatment  methods:
      Reduction in toxicity, mobility or volume
      Preference for on-site remedies
      Cost effectiveness
      Permanent remedies
      Proven and innovative remedies
      ARARs and how RCRA action will effect viability of
      soil based biological treatment
      Competing technologies
        Incineration
        Chemical treatment
        Physical treatment  and  isolation

 2.  RCRA regulatory activities
      A. Land disposal ban
          Land and soils based treatment systems
          Disposal of treated soils from contained systems
          Air emissions from treatment facilities
          Allowable contaminant  migration in soils
          Best Demonstrated Available Technologies for land
          disposal of wastes
          Other related RCRA permitting requirements for
          construction,  operation  and  closure of  bio-
          remediation facilities
Case Studies
Superfund site remedy & RCRA lagoon closure in northern
Minnesota. Report includes the results of two full years of full
scale facility operations
Superfund site remedy in Washington  state
Site conditions
Timing
Goals
Other remedies besides source controls
Pilot and demonstration work
Accomplishments to date
Reductions and treatment efficiencies
Economics, capital and operating
Design & Operating Parameters

 1. Design Considerations
    a.  Waste Volumes
    b.  Waste Characteristics
         physical composition
         organics
         metals
         salts
         nutrients
         pH
    c.  Site Characteristics
         topography/slope
         soil texture
         soil permeability
         soil pH
         CEC
         nutrients
         climate
         depth to WT
    d.  Operational Characteristics
         oil loading rates
         frequency  of application
         hydraulic loading rate
         solids loading rate
         metal loading rate
         method of application
         depth of incorporation
         frequency  of cultivation
         nutrients
         moisture
         pH
    e.  Treatment Goals
         how clean is clean?
         duration of treatment
         degradation kinetics
         leachate testing

 2. Design and Construction Features
    a.  Site Preparation
         clearing/grubbing
         grading
         topsoil
         bulking agents
         fertilizer
         monitoring
    b.  Run-off/Run-on Control  Systems
         berms
         sumps
         25-year/24-hour  storm
         discharge options (MPDES/ WVHTP/recycle)
    c.  Irrigation Systems
         system configuration
         monitoring for irrigation requirements
         source of water
    d.  Lined  Systems
         reasons for installation
         materials (compatibility/strength)
         leachate collection system
         construction/welding
         apertures
         costs
    e.  Support Facilities
534    SEMINARS

-------
         garage
         operational  equipment  (tractors, loaders,  discs,
         rototillers, pumps, etc)
         health and safety equipment and facilities

3.   Operating Procedures
    a.  Cultivation
         Frequency
         Depth
         Number of passes
         Qualitative  criteria to be followed by the operator
    b.  Irrigation
         Frequency/monitoring
         quantity
    c.  Application of amendments
         bulking agents
         lime
    d.  Sampling and monitoring
         frequency
         locations
         techniques for soil sampling
         techniques for lysimeter sampling
         QA/QC
         airmonitoring
    e.  Inspections/Maintenance
    f.  Health and safety
         baseline medical
         training
         monitoring
         equipment
    g.  Record keeping
F.  Summary
    Bioremediation  . .  . what do we know now about its effec-
    tiveness and what additional information is needed for future
    applications and to satisfy future regulatory concerns?

     1. Decision-making regarding the use of bioremediation tech-
       niques:
         Waste characteristics
         Site characteristics
            Soils
            Climate
            Ground and surface wastes
            Topography and other engineering features
            Utilities and labor
            Neighbors
         Performance required . .  .  removal criteria or final
         concentration
         Treatability and treatment demonstration studies
         Field plot scale studies
         Regulatory requirements and other performance
         criteria
         Technical and cost effectiveness feasibility analysis
         ROD or permitting
         Design, construction, operation, closure

     2. Where and when to consider bioremediation

     3. One of the best laboratories is a site if we gather enough
       data and design the remedy appropriately

     4. The usefulness of bioremediation should not be limited
       by institutional  limitations (land ban)
                                                                                                      SEMINARS    535

-------
536    SEMINARS

-------
                                                      1987  Exhibitors
 3M, Hazardous Material
 Control Products
 3M Center Bldg, 223-6S-04
 St. Paul, MN 55144-1000
612-733-9233
 3M  Company—Hazardous  , Material  Control
 Products—3M Foams. 3M Foams have proven
 their suppression effectiveness during hazardous
 material clean-up that involves release of volatile
 organic compounds (VOC), air toxics, odors, and
 dust. These water based foams conform to the ter-
 rain and  last  hours, days, and even  weeks
 depending on the site requirements.
 ACZ Inc.
 P.O. Box 774018
 Steamboat Springs, CO. 80477
 ACZ Laboratory Division: Providing laboratory
 services specializing in RCRA Hazardous Waste,
 priority pollutants, groundwater, drinking water,
 and underground storage tank analysis. Offering
 excellent turn-around times and extensive quality
 control.
 AMO Pollution Services. Inc.
 6923 Ebenezer Road
 Baltimore, MD 21220
301-488-0800
ATEC Associates, Inc.
1300 Williams Drive, Suite A
Marietta, GA 30066-6299         404-427-9456
ATEC Associates, Inc. is a diversified engineering
firm with a staff of over 800 in 30 offices located
in principal cities throughout the United  States.
ATEC's Environmental Services Division provides
environmental consulting and remedial  action
services that include environmental audits,  RCRA
permitting,  remedial investigations/feasibility
studies, underground storage tank management,
asbestos surveys, landfill design, monitor/recovery
well design and installation, complete analytical
laboratory  capabilities, geophysical testing.
ATEC's  subsidiary,  Waste  Abatement  Tech-
nology, Inc. (WATEC), provides solid/hazardous
waste cleanup, in-situ biological  treatment and
asbestos abatement services.


Acres International Corporation
1000 Liberty Building
Buffalo, NY 14202-3592          716-853-7525
Acres  International  Corporation, an interna-
tionally known consulting engineering and  project
management firm, provides services to the solid
and hazardous waste industry including:  hydro-
geological investigations, ground water monitoring
and evaluations; design of treatment systems and
remediation  programs;  and  facility  closure
planning.
Advanced Environmental
Technology Corp.
1603-1605 Goldmine Road
Flanders, NJ 07836
                                                 201-347-711
                 AETC is a full service company offering pack-
                 aging, transportation and disposal for virtually all
                 types of drum and  bulk chemical wastes. Lab
                 chemical packaging, hazardous waste  and site
                 cleanup, reactive  and  explosive  disposal and
                 24 hour emergency  responses are also part of
                 AETC services.  Branch facilities  are located in
                 New Jersey,  New York, Pennsylvania, Massa-
                 chusetts and  North Carolina.

                 Agency for Toxic  Substances
                 1600 Clifton Road (Mailstop 28)
303-879-6260     Atlanta, GA 30333
                                404-454-4618
The Agency for Toxic Substances and Disease
Registry (ATSDR) is a part of the public health
service. It was created by Congress to implement
the health related sections of laws that protect the
public from hazardsous wastes or environmental
spills of hazardous substances. The two program
offices of ATSDR  are the Office  of External
Affairs, and the Office of Health Assessment.

Alcoa Separations Technology Div.
181 Thornhill Road
                                                 Warrendale, PA 15069
                                                                                412-772-0080
                Alcoa carries a line of products ideal for Super-
                fund site cleanup. The products range from
                aluminas  tailor-made  for  aqueous  metal  ion
                removal to  unique inorganic membrane units
                which are chemically inert. A new high capacity
                mixed-metal oxide adsorbent has been developed
                which is highly selective for the environmentally
                dangerous anions such as selenium, arsenic and
                chromate.

                Alliance Technologies Corporation
                213 Burlington Road
                Bedford, MA  01730             617-275-9000
                RCRA/CERCLA  related remedial engineering,
                field sampling, laboratory analysis, and ground-
                water monitoring and modeling. Mobile hazardous
                waste laboratory. Site investigations and air toxics
                monitoring. Complete RCRA permit application
                assistance. Incinerator trial burns. Closure and
                Post Closure Plans. Registered Engineers, Geolo-
                gists, and Industrial Hygienists. AIHA Certified
                Laboratory.

                 American Health  & Safety
                 6250 Nesbitt Road
                 Madison, WI 53719              800-522-7554
                American Health & Safety, Inc. is a nationwide
                industrial safety supply house featuring a full line
                of on-the-job safety products. We have over 3,000
                line items which are distributed throughout the
                safety industry, including asbestos, laboratory and
food industries. American Health & Safety special-
izes in the hazardous materials and toxic waste dis-
posal  fields. We will  be  displaying gloves,
respirators, coveralls, boots, safety glasses, tape,
shovels, instrumentation and first aid used heavi-
ly in the hazardous materials industry.
                                                American NuKEM Corporation
                                                454 S. Anderson Road, BTC532
                                                Rock Hill, SC 29730
                               803-329-9690
Chemical and radioactive waste process systems
engineering and management services. Clean-up,
removal and treatment of hazardous materials;
certified analytical laboratory and field services are
provided   through  subsidiary  companies:
AnalytiKEM (609) 751-1122; a full service environ-
mental laboratory specializing in sample collection
of environmental and hazardous waste samples;
ThermalKEM  (803) 329-9690;  Incineration  of
hazardous  toxic materials;  CyanoKEM (313)
933-1850; Chemical treatment of cyanides, sul-
fides, corrosives, toxics, and plating wastes and
a wide variety of inorganic wastes; An Emergency
Response TreatmentServicesTeam (803) 329-9690;
Coordinate all of these integrated services.
Analytical & Environmental Testing
1717 Seabord Drive
                                                                                                Baton Rouge, LA 70810
                                                                                                504-769-1930
                                                Analytical and Environmental Testing, Inc. is an
                                                independent testing laboratory  and consulting
                                                service with special emphasis on environmental and
                                                chemical analysis. Our established goal is to offer
                                                industry and business an opportunity to have ana-
                                                lytical problems characterized and solved effec-
                                                tively.  Laboratories in Baton Rouge,  LA and
                                                Mobile, AL allow interfacing with industries all
                                                across the United States. Analytical testing services
                                                include: water/wastewater evaluations, hazardous
                                                waste characterization,  soil/sediment chemistry,
                                                product evaluations, particle identification, bioas-
                                                say/biotoxicity evaluations.

                                                Aqua Magnetics International, Inc.
                                                1010 A Park Court
                                                Safety Harbor, FL 34695         813-447-2575
                                                Magnetic Power Units for treating of fluids, scale,
                                                salt, Paraffin and carbon in water, gases and fuels.
                                                Aqua Tech Environmental
                                                Consultants, Inc.
                                                P.O. Box 436, 181 S. Main  Street
                                                Marion, OH 43302-0436         614-382-5991
                                                Aqua  Tech  Environmental  Consultants,
                                                Inc.(ATEC) is a full service environmental testing
                                                laboratory providing  water, waste water, and
                                                hazardous waste analysis to  government and  in-
                                                dustry. ATEC's state-of-the-art instrumentation
                                                includes AA, ICP, GC, HPLC,  IR, and GC/MS
                                                                                                                      EXHIBITORS    537

-------
equipment. Rapid turn around and electronic data
transfer are two of ATEC's special services avail-
able to our customers.
Art's Manufacturing & Supply
Harrison at Oregon Trail
American Falls, ID 83211
208-226-2017
AMS, the leader in hand-operated soil sampling
equipment for over 40 years, is now the leader in
sampling equipment for the hazardous waste in-
dustry. Stop by Booth 214 and see the new patent-
pending soil recovery  auger and our full line of
completely stainless steel soil sampling equipment.

Association of Engineering Geologists
62 King Phillip Road
Sudbury,  MA 01776             617-443-4639
The Professional Society which  represents prac-
ticing  engineering  geologists  and  geological
engineers involved with practice related to surface
and subsurface water and contaminates; waste
management;  aggregate  production;  geologic
hazards; and the evaluation, planning, design, con-
struction, maintenance and operation of fixed
engineering projects.
 Autumn Industries, Inc.
 518 Perkins-Jones Road
 Warren, OH 44483
216-372-5002
Autumn Industries, Inc. is a licensed hazardous
material and waste transporter, specializing in bulk
solids. Skilled management and trained personnel
possess the ability to provide 200 units to serve
accounts in the Midwest and Northeastern states.
Autumn Industries, Inc. is fast becoming a major
force in  the hazardous transportation Field.


BCM Engineer!!
1 Plymouth Meeting Mall
Plymouth Meeting, PA 19462    215-825-3800
Quality engineering in Hazardous Waste Manage-
ment and Control: Groundwater Studies, Geo-
physical Surveys, Remedial Design Engineering,
Superfund Site Investigations, Facility Permitting,
Closure Plans, Real Estate Contamination Assess-
ments, Asbestos Surveys, Analytical Services.
 BES Environmental Specialists, Inc.
 82-86 Boston Hill Road
 Larksville, PA 18651              717-779-5317
 Complete Environmental Management and Service
 Company.
Battelle Pacific Northwest Labs
P.O.Box 999
Richland, WA 99352
509-375-2867
 Battelle Pacific Northwest Laboratories offers a
 wide variety of R&D and technical application
 services including site characterization and assess-
 ment for active and inactive sites, health effects
 assessment, and process control and remediation
 technologies. Battelle offers advanced technology
 couples with a cost-effective,  multi-disciplinary
 approach for solving waste-site cleanup problems.
 Bergen Barrel & Drum Company
 43-45 O'Brien Street
 Kearny, NJ 07032                201-998-3500
 Plastic drums both open and closed head, for the
 safe packaging and transport of regulated and non-
 regulated chemicals. These drums are excellent for
 incineration and disposal of hazardous wastes both
 liquid and solid. A complete line of tanks and car-
 boys can also be seen at our booth.

 538     EXHIBITORS
Bloprocess Engineering, Inc.
P.O Box 5936
Wilmington, DE 19808-0936      302-995-6010
Company provides engineering services for bio-
remediation of hazardous waste sites. Emphasis
is on optimization of site conditions to facilitate
and  promote  the growth  of aerobic microor-
ganisms capable of degrading site materials in-silu.
Bio-cleanup methodologies include slurry reactors,
landfarming and surface biotreaters for decon-
tamination of  groundwater, sludges and soils.

Bird Environmental Systemf, Inc.
100 Neponset  Street
South Walpole, MA  02071        617-668-0400
Manufacturer of 5 models of mobile dewatering,
oil recovery and pretreatment systems. Each is
based on different separation  technology,  in-
cluding decanter and disc centrifuges, belt filter
and plate and frame filler presses. Two facilities
support sales  and  marketing of these systems,
along with special services provided in permitting,
laboratory analysis and technical assistance.

Black &  Vealcn
1500 Meadow Lake Parkway
Kansas City, MO 64114          913-339-2000
A nationwide consulting firm providing complete
engineering services pertaining to hazardous waste
management  including  remedial investigations,
feasibility studies, design of remedial actions, im-
plementation oversight, RCRA services, regulatory
and  permit support,  and litigation assistance.
Other specialties include waste treatment, waste-
to-energy systems,  public  health evaluations,
facility closure services, training, and environmen-
tal audits.
                 Brown A Caldwell Laboratories
                 373 South Fairoaks Avenue
                 Pasadena. CA 91105
                                818-795-7553
                 Brown & Caldwell Laboratories provides physi-
                 cal, chemical and biological analyses in two fully
                 equipped facilities, which were among the first cer-
                 tified  under a  1986 California law  that set new
                 standards for hazardous waste analyses. Services
                 include on-site sampling, monitoring and analy-
                 sis of water, groundwater, soil, wastes, and air;
                 trace  organic and  inorganic chemical  analyses;
                 microbiology; toxirity bioassay; and radiochemi-
                 cal analyses.
                 Brucker Instruments, Inc.
                 Manning Park
                 Billericia, MA 01821
                                                                                 617-667-9580
The  Bruker  MOBILE  ENVIRONMENTAL
MONITOR: A fully mobile GC-mass spectrometer
for on-site analysis of organic pollutants.

Bryson Industrial  Services, Inc.
411  Burton Road
Lexington, SC 29072             803-845-7027
Bryson Industrial  Services,  Inc. is a hazardous
waste management company. We provide consul-
tation and management services to customers on
methods of reducing, handling and disposing of
their hazardous waste.  In addition we provide
secure permitted transportation and fully trained
and experienced project teams for on-site service
needs.

Bureau of National  Affairs, Inc.
 1231 25th Street,  NW Room S-501
Washington, DC 20037           202^*52-4452
BNA publishes regulatory, legal,  and  working
guides providing the latest information concerning
the  manufacture,  transportation, safe handling,
and disposal of hazardous materials.
C-E Environmental Inc.
A Subsidiary of Combustion Engtotcrlng inc.
1 Becker Farm Road
Roseland, NJ 07068             201-992-2323
C-E Environmental Inc., a subsidiary of Combus-
tion Engineering Inc., provides fuD service environ-
mental management solution* in meeting the needi
of industry and the public sector. Included in this
offering are consulting and monitoring, remedial
action and field services, and recently acquired an
environmental engineering firm, E.C. Jordan Co.,
to further expand this offering and has acquired
exclusive North American rights to the Widmer
& Ernst slagging rotary kirn incineration technol-
ogy. The license will complement other technolo-
gy offerings,  including ashing kilns, chemical and
biological systems, and a recent marketing agree-
ment with  Keller/Dorr Oliver for fluidized bed
technology.
                                                                 CDF Corporation
                                                                 100 Enterprise Drive
                                                                 Marshfield, MA 02050
                                 617-837-2823
                                                                 CDF Corporation "The Liner People" offers the
                                                                 most comprehensive array of disposable plastic
                                                                 liners for drums, pails and tanks.
CECOS International
2321  Kenmore Avenue
Buffalo, NY 14207
                                                                                                                                  716-873-4200
                                                                 CECOS International, Inc., is a company specia-
                                                                 lizing in the treatment and disposal of hazardous
                                                                 chemical waste. CECOS makes these services avail-
                                                                 able to industry through a network of regional
                                                                 treatment centers  across the United States and
                                                                 Puerto Rico. CECOS offers specialized hazardous
                                                                 waste capabilities, research and analytical and con-
                                                                 sulting services.
 CF Systems Corporation
 25 Acorn Park
 Cambridge, MA 02140
                                                                                                  617-492-1631
 The CFS extraction process is a solvent extraction
 technique which, instead of using a typical solvent
 such as methylene chloride, touene or hexane, uses
 a  liquefied gas such as CO2, propane, or other
 light hydrocarbon gas. These solvents have high
 solubilities for most organic compounds that are
 listed as hazardous. They are also inexpensive,
 non-toxic and can be relatively easily separated
 from the extracted compounds. These properties.
 together with CF Systems proprietary equipment
 design, lead to a highly effective broadly applica-
 ble process with low operating costs. In general,
 the  CF Systems units  can extract  over 99% of
 liquid hydrocarbons  from  liquids and sludges
 having solids  and hydrocarbons content in any
 ratio.
 CH2M Hill
 Post Office Box 4400
 Reston. VA 22090
                                                                                                  703-471-1441
                                                                  CH2M HILL is a consulting engineering firm with
                                                                  over 40 offices throughout the world. The firm is
                                                                  a full service architect/engineering firm and pro-
                                                                  vides complete environmental engineering and en-
                                                                  vironmental services to industry and government.
                                                                  CH2M  HILL  is  ranked by Engineering News
                                                                  Record as the 3rd largest A/E firm in the United
                                                                  States.

                                                                  Calgon Carbon Corporation
                                                                  Post Office Box 717
                                                                  Pittsburgh. PA 15230            412-787-6700
                                                                  Calgon Carbon Corporation supplies activated car-
                                                                  bon products, systems, and services and airstrip-
                                                                  pers to remove soluble organic compounds from
                                                                  contaminated  groundwater,  surface water  or
                                                                  wastewater.

-------
California District
Attorneys Association
1130K Street, Suite 200
Sacramento, CA 95814           916-443-2017
At the California State Capitol, CDAA speaks to
the concerns of crime victims, the need for reform
in our justice system, and the importance of public
safety. CDAA's comprehensive publications cover
courtroom presentation, hazardous wastes  and
many other topics dealing with crime and punish-
ment. In  addition, CDAA holds one  to  two
statewide training seminars a month to assist prose-
cutors.
Cambridge Analytical Associates
1106 Commonwealth Avenue
Boston, MA 02215               617-232-2207
Cambridge Analytical Associates (CAA) provides
analytical laboratory, consulting, and bioremedi-
ation services. The Environmental Services Divi-
sion's Laboratory is certified in many states and
is an EPA approved Contract Laboratory (CLP)
for both organic and inorganic analyses. CAA's
Consulting Department provides data validation
and interpretation, site  investigations, data base
management, and chemometrics consulting.


Camp Dresser & McKee Inc. (CDM)
One Center Plaza
Boston, MA 02108               617-742-5151
CDM is a leading manager of large, complex
projects aimed at solving domestic and internation-
al environmental problems. The firm melds tra-
ditional engineering capabilities with innovative
technological approaches to achieve permanent
solutions to  today's hazardous waste problems.
CDM serves both government and industry areas
including design and construction, hazardous and
solid waste management,  water resources, and
regulatory compliance.  Our expertise also incor-
porates training, policy development, and expert
systems.
 CarbonAir Services
 Post Office Box 5117
 Hopkins, MN 55343
612-935-1844
 CarbonAir Services provides treatment design and
 installation  for  system removal of  dissolved
 organic or inorganic contaminants in groundwater,
 surface water, or air streams. Treatment alterna-
 tives include carbon adsorption (liquid or vapor
 phase), packed column airstripping,  oil/water
 separation (membrane or physio-chemical), heavy
 metals precipitation, and ancillary equipment for
 turbidity removal,  solids dewatering, etc.
 Carbtrol Corporation
 39 Riverside Avenue
 West Port, CT 06880
203-226-5642
Carnow, Conibear & Associates, Ltd.
333 W. Wacker Drive
Chicago, IL 60606               312-782-4486
Carnow, Conibear  and Associates, Ltd., with
offices in Chicago, Washington,  DC, and Los
Angeles, provides consulting services to address
occupational and environmental health concerns
associated with hazardous waste. CCA evaluates
health effects due to exposure to hazardous sub-
stances, designs and implements medical surveil-
lance programs, provides medical exams, trains
employees, (24 hour RCRA, 40 hour CERCLA
courses available), develops and implements site
safety and health plans, offers industrial hygiene
services.
                 The Center for Hazardous
                 Waste Management
                 10 West 35th Street
                 Chicago, IL 60616               312-567-4250
                 The Center for  Hazardous Waste Management
                 provides an integrated capability to deal with all
                 hazardous waste issues.  Services include: basic
                 R&D covering pollution control and waste reduc-
                 tion; quality assessment of site remediation activi-
                 ties; education training;  environmental law and
                 policy; environmental compliance; site assess-
                 ments; SARA, Title III;  and custom software.


                 Chem-Met Services
                 18550 Allen Road
                 Wyandotte, MI 48183            313-282-9250
                 Waste stabilization by fixation.. .21 years serving
                 industry. Processing and  transportation covering
                 36 states and two provinces. Chem-Met is a cost
                 effective, efficient way to handle industries waste
                 problems.


                 Chemfix Technologies, Inc.
                 2424 Edenborn Avenue,  Suite 620
                 Metairie, LA 70001               504-831-3600
                 Chemfix  Technologies,  Inc.  (CTI)  offers  the
                 patented CHEMFIXฎ process for chemical fixa-
                 tion/stabilization of  both hazardous  and non-
                 hazardous liquids and sludges. Complete mobile
                 services are offered, as well as fixed plant facili-
                 ties for continuous generation waste streams. CTI
                 services  include site assessments, waste stream
                 characterization and permitting support.
                 Chemical Waste Management, Inc.
                 3003 Butterfield Road
                 Oak Brook,  IL 60521
                                                 312-654-8800
                 Chemical Waste Management, Inc. is the world's
                 largest company involved in the analysis, trans-
                 portation, treatment and disposal of hazardous
                 wastes. The company's ENRAC division special-
                 izes in remedial cleanup projects  and offers on-
                 site treatment as well as off-site  treatment and
                 disposal.
Chromanetics
709 North Blackhorse Pike
Williamstown, NJ 08094
                                                 609-728-6316
                 Complete catalogue of laboratory and field sup-
                 plies for the environmental firm. Product  lines
                 include sampling devices, volumetric glassware,
                 extraction glassware,  (AA, IR, UV, GC, HPLC
                 instruments   accessories), analytical standards,
                 glass and  plastic containers,  lab furniture and
                 numerous other products. The company also pro-
                 vides a complete line of bottles, sampling devices
                 and safety equipment for the most discriminate
                 field user.
                                                 Clayton Environmental Consultants
                                                 22345 Roethel Drive
                 Novi, MI 48050
                                                 313-344-1770
                 Environmental Engineering Services.  • Environ-
                 mental Risk  assessment  and corrective action
                 strategies. • Point source and ambient air quality
                 studies. • Regulatory agency liaison. •  Under-
                 ground storage tank testing and  management.
                 • Hazardous waste disposal, storage, handling,
                 and training programs.  • Geological and hydro-
                 logic evaluation.  • Chemical emergency response
                 programs. • Comprehensive surveys, audits, and
                 program development.  •  State and Federal permit
                 application  preparation  and  negotiation.  •
                 Groundwater and waste-water studies.  •  Fugitive
                 emissions inventories and odor studies.
                                                Clean Sites, Inc.
                                                1199 North Fairfax Street, #400
                                                Alexandria, VA 22314
                                703-683-8522
                                                CSI is the independent,  non-profit organization
                                                dedicated to speeding the cleanup of hazardous
                                                waste sites.  It helps parties responsible for sites to
                                                organize themselves and  allocate costs, as well as
                                                assisting responsible  parties  and  governments
                                                achieve   legal  settlements  for cleanups. CSI
                                                manages and  reviews  RI/FS's and manages
                                                cleanups.
                                                Clean Sites, Inc.
                                                1199 North Fairfax Street, #400
                                                Alexandria, VA 22314           703-683-8522
                                                CSI is the independent,  non-profit organization
                                                dedicated to speeding the cleanup of hazardous
                                                waste sites.  It helps parties responsible for sites to
                                                organize themselves and  allocate costs, as well as
                                                assisting responsible  parties  and  governments
                                                achieve   legal  settlements  for cleanups. CSI
                                                manages and  reviews  RI/FS's and manages
                                                cleanups.
                                                CompuChem Laboratories, Inc.
                                                3308 Chapel Hill/Nelson Highway
                                                Research Triangle Park, NC
                                                27709
                                                                                800-833-5097
CompuChem Laboratories is a full service and
CLP approved laboratory. CompuChem is offer-
ing Third Edition SW-826 RCRA Analyses, along
with  Superfund Analyses,  Priority Pollutant
Analyses,  and Dioxin Analyses. CompuChem is
introducing the Environmental Site Profile (ESP)
System, a proprietary data management system.
ESP provides on-line access to your lab test results,
plus the capability for  flexible analysis and the
presentation of downloaded  data.
Continental  Vanguard
204 Harding Avenue
Bellmawr, NJ 08031              609-931-0950

Continuing Engineering Education  1905-FTO
George Washington University
801 22nd Street, NW                   '
Washington, DC 20052           202-994-7406
Corroon & Black
330 E. Kilbourn Ave
1 Plaza E. 1400
Milwaukee,  WI 53202            414-271-9800
If you are involved or thinking of becoming in-
volved with remedial/hazardous waste work, you
face new and unusual risks. Corroon & Black, one
of America's  largest construction industry in-
surance brokers, has pioneered much of the ex-
pertise you need. By combining risk management,
commercial insurance (including pollution liability
coverage) and indemnification, risks can be iden-
tified and controlled. Corroon & Black has an in-
surance  program  specifically for hazardous
waste/remedial action contractors. It includes pol-
lution legal  liability, general liability,  business
automobile and property coverages.

Crown Andersen Inc.
306 Dividend Drive
Peachtree City, GA 30269         404-997-2000
Crown Andersen Inc. engineers, manufactures and
installs complete  hazardous waste  incineration
plants and handles permitting of such plants in ac-
cordance with RCRA regulations. The company
also  manufactures  non-metallic,  high  integrity
tanks, containers, and storage vaults. These con-
tainers are used for hazardous waste materials,
radioactive waste materials, and corrosive chemi-
cals.  The company  offers services ranging from
waste disposal process development through com-
plete plant design, including financing arrange-
ments. Company processes  are used to  handle
liquid, gaseous, solid, and sludge wastes. The com-
pany also  offers  equipment  and  services  for
                                                                                                                      EXHIBITORS     539

-------
 hazardous waste incineration plant modernization.
 Working models of some of the company's equip-
 ment will be displayed.
 DART/Lodestar
 61 Railroad Street
 P.O. Box 89
 Canfield, OH 44406             216-533-9841
 The DART consortium offers Dan Trucking Com-
 pany as one of the nation's largest hazardous waste
 transporters, providing over 450 vehicles such as
 semi-dumps, rolloffs, bulk pneumatics, box vans
 and flatbeds. Dart is also  involved in joint ven-
 ture partnership known as Lodestar, which has
 begun manufacturing totally aluminum roUoff
 trailers and boxes to handle greater payloads. Dan
 Services, Inc. provides environmental services/pro-
 ducts,  handling private  company's  live waste
 stream needs, with disposal,  reprocessing  or
 recycling options; sorbent, solidification and neu-
 tralization agents for product spills or stabilizing
 wastes;  environmental software packages and
 regulatory-mandated training.
 DECI, Inc.
 11 Penn Plaza
 New York, NY 10001            212-216-7070
 Dravo Engineering Companies, Inc., engineering,
 design, and construction management services for
 hazardous waste  investigations,  treatment, and
 disposal.
 DataChem, Inc.
 520 Wakara Way
 Salt Lake City, UT 84108        801-583-3600
 Full service environmental and industrial hygiene
 laboratory.
 Dexsil Corporation
 P.O. Box 6556
 Hamden, CT 06517              203-288-3509
 Dexsil Corporation exhibiting portable, quick,
 accurate and easy to use testing kits. The Clor-N-
 OUTM PCB Screening Kit  for checking trans-
 former oil for PCB content, and our newest kit,
 called Clor-D-Tect™ checks all types of waste oil
 for hazardous chlorinated solvents. A full service
 analytical laboratory and more.
 Dobrmmo
 3240 Scott Boulevard
 Santa Clara, CA 95054           408-727-6000
 Total organic carbon and total  organic halide
 analyzers for water, total halide analyzers for used
 oil and  solvents, trace  sulfur  and  nitrogen
 analyzers.

 Du Ponl Company/
 Environmental Services
 External Affairs Dept., Room NA235
 Wilmington, DE 19898           302-774-2692
 Through its Environmental Services operation, Du
 Pont treats wastewater and contaminated equip-
 ment on a contract basis at its EPA-permitted
 Chambers Works facilities, Deepwater, NJ. The
 40-million-gallon-a-day Wastewater  Treatment
 Plant destroys  industrial  wastewater  chemicals
 through the patented Powdered Activated Carbon
 Treatment (PACT) process. The Thermal Decon-
 tamination Unit removes organic chemical  re-
 siduals from equipment through high-temperature
 treatment.

 DuPont/Envlronmental
Management Services
External Affairs Dept., Room N252I
Wilmington, DE 19898           302-774-2692

DuPont Environmental  Management  Services
offers programs and consulting services tailored
 to supplement and enhance our client's capabili-
 ties to meet their environmental responsibilities.
 Our Focus is on developing efficient, cost effec-
 tive and practical solutions to real world environ-

 540    EXHIBITORS
mental problems.
Dunn Geosclence Corporation
12 Metro Park  Road
Albany, NY 12205               518-458-1313
Dunn Geoscience's geotechnical consulting at
CERCLA/SARA and RCRA sites  ranges from
modeling pollutant transport to characterizing the
hydrogeologic environment for remedial design.
More than a quarter century of excellence in
investigation and documentation of subsurface
environments is the firm's foundation for substan-
tial involvement in hazardous waste remediation.
Dynamac Corporation
11140 Rockville Pike
Rockville, MD  20852             301-468-2500
Dynamac Corporation, an engineering and scien-
tific services firm, has performed approximately
600 hazardous waste/materials projects since its
inception in 1970. We can solve your problems in
ares such as waste site investigation, evaluation and
remediation; risk/endangerment assessment; oc-
cupational safety and health; compliance auditing;
RCRA permitting; and corrective action.

Dyniiherm Analytic Instruments Inc.
P.O. Box  159
Kelton-Jennersville Rd.
Kelton. PA 19346               215-869-8702
Exhibiting laboratory equipment for air, water and
soil analysis. Concentrating instruments to add to
GC/GC-MS to  improve productivity and sensi-
tivity levels. Techniques include: purge/trap, head-
space and steam distillation for water, soil and
hazardous materials analysis,  thermal desorption
of air monitoring and sample collection tubes.
Low-cost manual  and automated  systems
available.
EBASCO Services Inc.
160 Chubb Avenue
Lindhurst, NJ 07071            201-460-6485
EBASCO Services Incorporated is a full-service
consulting, engineering and construction company
offering toxic materials, hazardous and mixed
waste management service. These services include
remedial investigations, feasibility studies, chem-
ical and environmental engineering,  geotechnical
and geohydrological consulting, project manage-
ment and construction, environmental planning,
consulting and design, quality assessment, finan-
cial and other specialties.

ECOFLO
2750 Patterson  Street
Greensboro, NC 27404           301-773-9500
ECOFLO offers complete "turnkey" hazardous
waste disposal services. From waste characteriza-
tion to disposal  certification,  ECOFLO manages
most chemical wastes. ECOFLO owns and oper-
ates a waste pretreatmem facility in  Greensboro,
NC and provides field service and transportation
throughout the  eastern U.S.

EMTECH Environmental Services, Inc.
1 Summit Avenue, Suite 206
Fort Worth, TX 76102           817-332-5481
EMTECH Environmental Services, Inc., is a sin-
gularity focused  corporation committed to provide
environmental services to both public and private
sector clients. EMTECH's activities are in the area
of environmental consulting services, and provides
full scope of remedial action services, combining
to offer single point responsibility to our clients.
ENRECO, Inc.
P.O. Box 9617
Amarillo, TX 79105             806-379-6424
ENRECO, Inc.  is the leader in stabilization, treat-
ment, and disposal  technologies. Years of  field
experience have evolved proven methods, techno-
logies, and equipment for the solidification and
fixation of waste products. ENRECO is supported
by  the analytical  and engineering services of
ENRECO  Laboratories  and  ENRECO  En-
gineering.

ENSECO, Inc.
4955 Yarrow Street
Arvada, CO 80002               303-421-661J
ENSECO, Inc. provides nationwide  laboratory
services from facilities in Boston, MA; Berkeley
Heights.  NJ;  Richmond,  VA; Houston, TX;
Denver, CO; and  Sacramento, CA. ENSECO
specializes in  providing consultative analytical
chemistry and aquatic toxicology services to solve
environmental problems for industrial clients and
governmental agencies. The corporation also has
capabilities in industrial hygiene, pharmaceutical
chemistry, gas  analysis and  industrial problem
solving.
ERM Computer Services,  Inc.
999 West Chester Pike
West  Chester. PA 19382         215-696-9110
ERM Computer Services. Inc. provides informa-
tion services related to environmental engineering
through  a software  system  called  ENFLEX.
ENFLEX DATA is a comprehensive environmen-
tal  information management service providing
reports and graphs for field operations personnel
along with summary reports for corporate level
staff.  ENFLEX INFO is a  powerful information
system that provides direct access to the full text
of current federal and state environmental regu-
lations using a personal computer.
ERM  Group. The
999 West Chester Pike
West  Chester, PA 19382          215-696-9110
The ERM Group is a  full  service environmental
engineering/management consulting firm. With 26
affiliate offices throughout North America, we
provide technical services to solve any environmen-
tal  facility  or regulatory management  need:
Hazardous/solid waste management, industri-
al/municipal water and wastewater  treatment,
environmental management consulting, biological
investigations, air pollution control, hydrogeolog-
ical investigations, underground tank and com-
puterized data base management.

Eagle-Picher Environmental Services
200 9th Avenue. NE
Miami, OK 74354                800-331-3144
Eagle-Picher  Environmental Services is a fufl
service analytical laboratory providing a wide spec-
trum  of analytical services. Eagle-Picher is the
western region bottle repository contractor for the
U.S. EPA contract lab laboratory program. It also
has a  full scale organic synthesis laboratory and
provides  organic standards. The field services
group is capable of providing complete sampling
and analytical services. The laboratory provides
low level  detection of environmental analysis
including  dioxins  and other  hazardous  list
compounds.

Earth Resources Corporation
Post Office  Box 616961
Orlando, FL 32861-6961          305-295-8848
Earth Resources  Corporation (ERC) is  a  full
service hazardous  materials management firm
specializing in the containment, treatment, and
removal of all types of hazardous materials. ERC
has a highly trained professional and technical staff
experienced in the design and implementation of
innovative solutions to today's waste problems.
ERC has recently introduced the Cylinder Recon-
tainerization  Vessel, designed  to  safely handle
compressed gas cylinders and transfer their con-
tents to secure containers. Remedial action capa-
bilities  also   include  groundwater,  facilities
decontamination, and soil removal.

-------
Earth Technology Corporation, The
3777 Long Beach Boulevard
Long Beach, CA 90807          213-595-6611
The Earth Technology Corporation offers com-
prehensive hazardous waste management services.
Those which we are most frequently requested to
provide include: Environmental auditing/compli-
ance  assessment, hazardous  waste permitting,
remedial  investigations,   remedial/corrective
action, design  and engineering, geotechnical in-
vestigations, waste stream reduction and recovery,
facility closure and laboratory and special techni-
cal services.

Eastern Chemical Waste Systems
1101  14th Street, NW #620
Washington, DC 20005          202-289-5490
Eastern Chemical  Waste  Systems  (ECWS), a
division of The Soresi Chemical Group, Inc., is
a complete hazardous waste management company
offering a wide range of capabilities and services
designed to help you comply  with all of your local,
state  and federal requirements for the transpor-
tation and disposal of your hazardous and toxic
chemical wastes. ECWS is a full service operation,
presently serving over 500 customers in the United
States and Puerto Rico. ECWS's major services
include Hazardous Waste and Laboratory Chem-
ical  Disposal,  Waste  Analysis, Chemical Spill
Clean-Up,  Plant Closure Clean-Up and Decon-
tamination, Transportation, Consulting  Services,
Emergency Response,  Incineration, PCB Clean-
Up and Remedial Action. In addition, ECWS has
an Infectious Waste Division, specializing in the
removal and disposal of biohazardous waste, and
a sister company, Volta Environmental Services,
Inc.,  specializing in turn-key PCB equipment
replacement.

Ecology and Environment,  Inc.
Post Office Box D
Buffalo, NY 14225               716-632-4491
Ecology and Environment, Inc. provides complete
hazardous waste engineering services—site inves-
tigations,  remedial  plans  and  specifications,
hydrogeological studies; air, water and  ground-
water monitoring, analytical laboratory  services;
spill  emergency  response;  asbestos   removal
management; hazards and risks analyses; under-
ground storage tank management; environmental
impact assessments.

Ecova Corporation
15555 NE 33rd Street
Redmond,  WA 98052            206-882-4364
Ecova Corporation  designs,  constructs  and
manages site-specific systems for the remediation
of hazardous and toxic waste  sites. Systems em-
ploy  state  of  the art biological, chemical and
physical technologies. Ecova services include site
assessment, feasibility, process design  and site
management.

Electrum Inc.
P.O.  Box 1
Fairfield, KY 40020              502-252-9944
Manufacturer  and supplier of water treatment
sorbents and chemicals.

Engineering-Science, Inc.
57 Executive Park
South Atlanta, GA  30329        404-325-0770
Engineering-Science (ES) is  a major, full service,
national and international  environmental engi-
neering firm  providing complete  services  in
hazardous waste management. With offices in 15
domestic locations, ES  is active in supporting in-
dustrial and  military  clients in all phases of
site/remedial investigations,  feasibility studies,
remedial   action  plan   preparation,   site
cleanup/closure  and post-closure activities.
Enviresponse, Inc.
110 South Orange Avenue
Livingston, NJ 07039            201-533-2385
Enviresponse, Inc., contractor for the U.S. EPA's
Environmental Emergency Response Unit in Edi-
son, New Jersey, offers a wide range of technolo-
gies and expertise relevant to the complex problems
surrounding the treatment of hazardous and toxic
waste.  Services range from  consulting to design
and construction. Enviresponse, Inc. is a subsidi-
ary of  the Foster Wheeler Corporation.

Envirite Field  Services, Inc.
600 West Germantown Pike #221
Plymouth Meeting,  PA 19462     215-825-8877
Envirite Field Services provides solidification/fixa-
tion services for organic and/or inorganic indus-
trial wastes. The company offers three proprietary
delivery systems to stabilize waste liquids, sludges
and contaminated soils with  selected additives-the
VR/STM system for  low-range solids, the PF-5TM
system for mid-range solids, and the HSSTM sys-
tem for high-range solids. The company also pro-
vides  dewatering services  using  mobile filter
presses.

EnviroQuip, Inc.
P.O. Box 31948
Palm Beach Gardens, FL
33410-7948                      305-694-4650
EnviroQuip "s fixation/stabilization program offers
low-cost low-risk EPA approved solutions to waste
disposal problems. This proven technology and ex-
pertise  is now available through EnviroQuip, Inc.,
a subsidiary of FPL  QualTec, Inc., who has suc-
cessfully tested and implemented a waste treatment
process for soils contaminated with PCB's, oils,
light hydrocarbons,  solvents and heavy metals.
Envirodyne Engineers, Inc.
12161 Lackland Road
St. Louis, MO 63146            314-434-6960
Environmental Engineering Services in Hazardous
Waste Management, Site Evaluations, Site Assess-
ment, Treatability Studies, treatment System De-
sign, Regulatory Liaison, Quick Response Teams,
Environmental Impact Assessment, Site Closure
Plans,  with an in-house High  Technology,  En-
vironmental Analytical Laboratory capable of per-
forming analysis of Water, Wastewater, Soils, Air,
and other Environmental Samples.

Environmental Compliance Systems
721 E. Lancaster Avenue
Downingtown, PA  19335         215-269-6731
ECS assists environmental or technical companies
with insurance, safety  and compliance  needs
through in-house expertise in environmental regu-
lation, technical risk management and insurance
underwriting.

Environmental Management News
225 North Newroad
Waco, TX 76714                817-776-9000
Environmental Management News magazine  is
dedicated to the management of air, water, waste
water,  pollution  and hazardous materials.  En-
vironmental Management News Action Pac offers
information on products and services for the con-
trol  of  industrial   pollution   and hazardous
materials.

Environmental Science & Engineering
P.O. Box ESE
Gainesville, FL 32602            904-332-3318
ESE offers complete  one-step environmental
services. Areas  of  service  include: Toxic  and
Hazardous Materials Control; Pollution Control
and Problem Solving; Surface and Ground Water
Monitoring; Underground Storage Tank Assess-
ment; Water and Waste Water Treatment Tech-
nology; Regulatory  Analysis, Permitting,  and
Compliance;  Source  Testing,  Ambient  Air
Monitoring and Air  Modeling; and Technology
Transfer.

Environmental Systems Company
333 Executive Court
Little Rock, AR 72205            501-223-4100
ENSCO operates one of the largest commercial
land based incinerators in the country at El Dora-
do, Arkansas serving the PCB and the RCRA mar-
kets. ENSCO Environmental Services is a full
service site remediation company including modu-
lar transportable incinerators for on-site appli-
cation. In addition site assessment, remediation,
closure and post-closure services are available. In
addition to modular incineration other technolo-
gies for on-site application are available through
ENSCO Environmental Services.
Environmental Technology, Inc.
Second and Maury Streets
Richmond, VA 23224
804-231-2232
Environmental Technology, Inc. is a diversified
environmental engineering firm that supplies geo-
technical consulting, environmental engineering,
and  municipal and industrial services. Services
include petroleum, chemical and waste tank evalu-
ation, decontamination, removal and disposal. We
also  provide site assessment, remedial action al-
ternatives, cost estimates and compliance monitor-
ing closure plans.

Envirosafe Services, Inc.
900 E. 8th Avenue, Suite 200
King of Prussia, PA 19406        215-962-0800
Envirosafe Services, Inc.-hazardous waste manage-
ment through  a  variety of subsidiaries: ACES,
Associated Chemical & Environmental Services-
provides remedial services, spill cleanups, biotech-
nology, groundwater filtration, site maintenance,
lagoon closures and underground storage tank ex-
cavation.  Fondessy Enterprises,  Inc.—disposal
services for hazardous/nonhazardous wastes, land-
farm capabilities. Envirosafe Services of Idaho,
Inc.—(ESII) secure disposal services for PCB-
contaminated materials.
Espey, Huston and Associates, Inc.
7700 Leesburg Pike
Falls Church, VA 22043
301-556-7770
A full spectrum engineering and environmental
consulting firm with capabilities in over 30 profes-
sional disciplines. Recent hazardous waste projects
include RI/FS/RD/RA on  highland  acid  pits
superfund site, design and permitting of disposal
facilities for Rollins Environmental Services, de-
sign of a central hazardous waste manangement
facility for a major utility company, and a broad
range of engineering services for superfund PRPs.
EH&A has offices in the Washington,  DC area,
throughout the southeast and southwest, and in
Europe and Mexico.
Essex Environmental
10500 Tube Drive
Hurst, TX 76053-7910
817-267-3319
Essex Environmental Industries introduces the
Enviropack, Generation II, Polyethylene overpack
salvage drum for the transporting of leaking pack-
ages of hazardous materials, including corrosive
metal drums: manufactured in compliance with re-
quirements as  set  forth in DOT E-9775  (A)
(DOT-34). Lightweight, safely contains the broad
range of materials including acids and corrosives.
Deformed  or damaged drums  are easily over-
packed, designed for stackable storage. Nestable
for freight savings. Call Essex Environmental at
1-800-423-8188  for additional information.
                                                                                                                     EXHIBITORS     541

-------
Farm Chemicals
37841 Euclid Avenue
Willoughby, OH 44094
216-942-2000
Floor-Daniel
3333 Michelson Drive
Irvine, CA 92730                714-975-5519
Fluor Daniel provides engineering, design, con-
struction, project management, project financing,
maintenance, and consulting services to a broad
range of clients worldwide. Fluor Daniel's Envi-
ronmental  Information Management group in-
cludes a spectrum of environmental and computer
science professionals, providing state-of-the-art en-
vironmental  management  software,  support
services, and management consulting.
Forestry Suppliers, Inc.
205 West Rankin Street
Jackson, MS 39204-0397
601-354-3565
You'll find more than our name implies at this ex-
hibit. Stop by to  see the finest in soil recovery
augers and probes, ground water/surface water
sampling and testing equipment, safety wear for
workers exposed to hazardous wastes, surveying/
engineering instruments and supplies—and more.
Sign up for our free, 464-page catalog.
The Foxboro Company
Dept. 120/N30-1E
Foxboro, MA 02035
617-549-2502
Instrumentation for providing quantitative and
qualitative information on hazardous waste and
spill site contaminants. These instruments can be
used at the waste site to locate areas of high vapor
concentration, to identify and determine concen-
tration levels of various organic compounds, and
to provide rapid, reliable screening/analysis for
volatile hydrocarbons in ground water samples.
GHR Engineering Assoc., Inc.
75 Tarkiln Hill Road
New Bedford, MA 02745
617-995-5136
GHR Engineering is a 150-person civil and en-
vironmental engineering firm with a strong focus
on Superfund site investigation and remedial en-
gineering; hydrogeologic assessment and computer
modelling; solid and hazardous waste regulatory
compliance  and  permitting; and  laboratory
analysis.
GKN Hayward Baker
1875 Mayfield Road
Odenton, MD 21113
301-551-8200
GKN Hayward Baker, affiliated with the world-
wide  GKN Keller Group, specializes in in-situ
ground modification. Our comprehensive range of
techniques includes slurry trench cut-off walls,
vibro systems.  Dynamic  Deep  Compaction,
grouting systems, and  anchors and mini-piles.
Grouting  techniques and cut-off walls are par-
ticularly appropriate to hazardous waste  con-
tainment.


GSX  Service*, Inc.
3527 Whiskey Bottom Road
Laurel,  MD 20707               800-638-4440
GSX is a full service company specializing in com-
plete  management, movement and disposal of
chemical waste. We have four strategically located
service centers, a secure chemical landfill and a
thermal oxidation incinerator. Our Emergency,
Remedial and Technical Response team can be
reached 24 hours a day and performs on-site clean-
up projects. GSX assists with all phases of com-
pliance  and documentation to satisfy State and
Federal regulations.
General Physics Corporation
10650 Hickory Ridge Road
Columbia, MD 21044-3698
301-964-6044
                General Physics Corporation is dedicated to tech-
                nology transfer serving industrial organizations
                and government agencies. Our services include
                clients in the nuclear, chemical, petrochemical,
                food, plastics, power and paper industries. We are
                assisting the U.S. Army in the disposal of chemi-
                cal warfare agents and the Department of Energy
                in the more efficient and safer operation of spe-
                cial nuclear materials facilities.
                General Research
                7655 Old Spring House Road
                McLean, VA 22102
                                703-893-5900
Comprehensive, flexible data management and
surveillance systems for the occupational health
and environmental professional.

GEO-CON, Inc.
Post Office Box 17380
Pittsburgh, PA 15235            412-244-8200
GEO-CON, Inc. is a full service construction com-
pany that specializes in the containment, removal
and stabilization of all types of hazardous wastes.
The firm has extensive background nationwide in
hazardous waste remedial work and the construc-
tion of new containment systems. It has all the
equipment a hazardous waste project calls for, in-
cluding state-of-the-art earth-moving equipment,
and advanced testing and monitoring equipment.

Geoscience Consultants, Ltd.
500 Copper NW,  #200
Alberquerque, NM 87102         505-842-0001
Environmental Services including ground water
modeling; ground water and soil cleanup; remedial
investigations; site cleanup oversight and manage-
ment;  environmental  auditing;  underground
storage tank services.
Greenhorne It O'Mara, Inc.
9001 Edmonston Road
Greenbelt,  MD 20770
                                                301-982-2800
Greenhorne & O'Mara, Inc., investigates and pro-
vides solutions for hazardous waste contamination
of land, water and air. Our multidiscipiinary staff
is knowledgeable in the requirements of RCRA,
CERCLA, SARA. TSCA, Clean Water Act, and
Clean Air Act. Applications include site charac-
terization and cleanup, permitting, environmen-
tal audits, and waste reduction.

Gronndwater Technology Inc.
220 Norwood Park South
Norwood, MA 02062             617-769-7600
Groundwater Technology, Inc.,  restores sites con-
taminated by hazardous  materials. Capabilities
include  environmental  assessments, chemical
recovery, aquifer restoration, monitoring, storage
systems management, engineering services, labora-
tory analysis, human health and risk assessment,
equipment manufacturing, computer modeling,
regulatory assistance and expert testimony. 40
offices and 700 employees.

Gundle Lining Syitems, Inc.
1340 E. Richey Road
Houston, TX 77073              713-443-8564
Gundle Lining Systems, Inc. is  recognized  as the
World Leader in the manufacture and installation
of High Density Polyethylene lining systems up to
100 mil thick.

H2M Group
575 Broad Hollow Road
Melville, NY 11747              516-756-8000
H2M Group is a multi-disciplined consulting firm
specializing in environmental, civil and structural
engineering, architecture, planning and environ-
mental science. The firm's full scope of profes-
sional services encompasses civil/site engineering,
community planning,  water resources manage-
ment,  wastewater  pollution  control,  indus-
trial/hazardous waste management and analytical
laboratory services.

HAZCO, Inc.
1347 E. Fourth Street
Dayton. OH 45402              513-222-1277
HAZCO is a national turn-key supplier of all the
health and safety equipment required to safely
respond to a remedial project or emergency haz-
mat incident. Services include site specific  PPE
packages delivered from stock; 24-hour access to
our Tech Service via  1-800-332-0435; rental of
HNUs, OVAs, instrumenu.and decon trailers; and
our Safety Network Approved  Purchasing  Plan
(SNAP). Our experience is your advantage.

HAZCON. Inc.
P.O.  Box 947
Katy. TX  77492                 713-934-4500
HAZCON is recognized as a leader in hazardous
waste solidification, to include wastes of and highly
organic nature.  Independent tests  show HAZ-
CON's advanced process superior to others, capa-
ble of toxicity reductions, controlled compressive
strengths, and little to no volume increase. HAZ-
CON, a safe, economic and solid solution.

HAZMAT Software Co.-AIA Corp, The
134 Middle Neck Road
Great Neck, NY 11021           516-829-5858
All types of HAZMAT Software, including the
EPA's oil & hazardous material technical assis-
tance data base (OHMTADS) and the U.S. Coast
Guard's, chemical response information system
(CHRIS). Each contains different information on
1400 and 1100 substances respectively. Informa-
tion is listed in approximately 118 data fields in
each. Abo M.S.D.S. software (The Commander).

HAZTECH, tec.
5280 Panola Industrial Boulevard
Decatur. GA 30035-4013         404-981-9332
HAZTECH. Inc. is the nation's largest, privately
owned hazardous waste cleanup contractor. HAZ-
TECH capabilities include facility decontamina-
tion  and demolition,  on-site water treatment,
lagoon closures, mobile incineration, ground-water
treatment and recovery, contaminated material ex-
cavation, and 24-hour emergency spill response.
Operation centers are located in Georgia,  Florida,
and New Jersey.

HDR Infrastructure, Inc.
8404 Indian Hills Drive
Omaha. NE 68114               402-399-1000
HDR Infrastructure, Inc. (HDR) specializes in in-
dustrial and  hazardous  waste management,
including remedial investigations and feasibility
studies  of hazardous  waste sites; design and
implementation of remedial action alternatives;
facility permitting; design of treatment, storage
and disposal facilities;  and closure/post-closure
planning. Industrial projects encompass the study,
design and implementation of industrial waste
treatment; ultra pure water, gas and chemical
systems; environmental permitting; process facil-
ities for high-tech industries.

HMCRI
9300 Columbia Boulevard
Silver Spring, MD 20910         301-587-9390
HMCRI is a unique, public, nonprofit, member-
ship organization  which  promotes the estab-
542     EXHIBITORS

-------
lishment and maintenance of a reasonable balance
between expanding 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 ex-
change information and experiences dealing with
hazardous materials. "Join HMCRI Today."
HNU Systems, Inc.
160 Charlemont Street
Newton, MA 02161
617-964-6690
Model  HW101  Hazardous Waste  Portable
Analyzer; IS101 Portable Intrinsically Safe Pho-
toionization Analyzer; PI101 Portable Photoioni-
zation  Analyzer;  Model  301 DP  Dedicated
Programmable Gas Chromatograph; Model 321
Compact Gas Chromatograph; Model 331 Com-
pact Dedicated G.C.
 Hansen WeatherPort
 P.O. Box 715
 Gunnison, CO 81236
303-641-0480
 Hansen WeatherPort was established in 1968 as
 a  family business at  Gunnison,  Colorado.
 WeatherPorts were orginally designed to shelter
 the exploration, construction, and drilling indus-
 tries with portability,  flexibility,  and  ease  of
 installation. WeatherPort remains a family busi-
 ness successful in diversifing and expanding its
 product line, who boasts a world wide track record
 for the fast and economical shelter for man,
 material and equipment.

 Harding Lawson & Associates
 7655 Redwood Blvd.
 P.O. Box 578
 Novato, CA 94948               415-892-0821
 Comprehensive hazardous waste management con-
 sulting services, including site characterization, risk
 assessment, ground-water investigation, geophysi-
 cal studies, remedial investigations and feasibility
 studies (RI/FS), remedial design, procurement,
 construction management, post-construction and
 post-closure monitoring, regulatory assistance and
 liaison, expert witness  and testimony. Over  40
 Superfund projects conducted.
 Hewlett-Packard Company
 2 Choke Cherry Road
 Rockville, MD 20850
301-921-6296
 Hewlett-Packard will exhibit a gas chromatogra-
 phy system for analysis of purgeable halocarbons
 and aromatics in wastewater and drinking water.
 The unit consists of a gas Chromatograph, purge
 and trap device, electrolytic conductivity detector
 and photoionization detector. Also we will exhibit
 a  gas chromatograph/mass  spectrometer  for
 environmental analyses.

 Hoyt Corporation
 251 Forge Road
 Westport, MA 02790             617-636-8811
 Manufacturer of Solvent Vapor Recovery/Air Pol-
 lution Control Equipment; Distillation Equipment;
 Odor Control Equipment; Liquid Purification
 Equipment.

 Hydro-Search, Inc.
 235 North Executive Drive, #309
 Brookfield, WI 53005            414-784-4588
 HYDROGEOLOGIC       CONSULTING
 SERVICES:  RI/FS—work plans, peer review
 existing work plans and reports, investigations,
 evaluate alternatives, implement remedial action
program;  ground-water  monitoring  and
assessments—review existing/proposed programs;
RCRA Compliance programs; landfill permitting,
siting, design and closure; site assessments for sale
or purchase; environmental audits.

I-Chem Research, Inc.
23787-F Eichler Street
Hayward, CA 94545             415-782-3905
A complete line of sample bottles, jars and vials,
supplied with teflon-lined closures attached and
available chemically cleaned and treated to exact
U.S. EPA protocols.  Also available  custom-
cleaned to your  exact  specifications.  I-Chem
Research is supplier of pre-cleaned sample bottles,
jars, and vials to the U.S.  Superfund program
nationwide. Also call 1-800-443-1689.

ICAIR, Life Systems, Inc.
24755 Highpoint Road
Cleveland, OH 44122            216-464-3291
ICAIR, Life Systems, Inc. (ICAIR) specializes in
human health and environmental effects assess-
ments (endangerment assessments,  public health,
toxicity, contamination, environmental  and risk
and impact assessments). ICAIR's uniqueness is
in its approach, integrating the experience of
ICAIR's core scientific and management staff with
the finest scientific and technical  minds in the
world.

ICF Technology
1850 K Street, NW, Suite 950
Washington, DC 20006           202-862-1100
ICF Technology—the scientific and engineering
subsidiary of  ICF incorporated—consults on
environmental and hazardous waste management
issues. We provide our clients with a full  range of
technical services, from site-specific investigations
and risk assessments to remedial design and con-
struction monitoring. The firm is headquartered
in Washington, DC.

ICM
163 SW Freeman
P.O. Box 803
Hillsboro, OR 97123             503-648-2014
Industrial Chemical Measurement is a diversified
designer and manufacturer of standard chemical
measurement instruments and standard or custom
control systems. ICM systems are used to measure,
record, and control processes based on values of
pH, conductivity, salinity, oxygen, temperature,
voltage, turbidity,  color  absorbance/trans-
mittance, or ions. Applications include controlling
waste streams, plating solutions, food processing,
environmental monitoring, sewage treatment, and
other laboratory and field measurements. Control
units provide single or dual outputs to control
pumps, motors, or alarms. 4-20MA, RS232, time
delays, and voltage outputs available.

IDS, Inc.
200 Monroe Turnpike
Monroe, CT 06468              203-261-4458
IDS, Inc., provides complete sludge handling. Our
capabilities include lagoon characterization testing,
pumping, dewatering with  belt press, chamber
press, (plate and frame) and centerfuge, cake dis-
posal and lagoon cleanup. In addition we specialize
in sludge stabilization/solidification and bioreduc-
tion of organic petroleum hydrocarbon liquid and
sludges.

INFORM
381 Park Avenue South
New York, NY 10016            212-6894040
INFORM is a non-profit research organization
that reports on practical actions for the preserva-
tion and conservation of natural resources. We
conduct seminars, provide  speakers  for confer-
ences and are frequently called upon to brief com-
mittees such as the U.S. Senate Environmental &
                                                                 Public Works, and the U.S. House of Represen-
                                                                 tatives Committee on Energy and Commerce. Our
                                                                 research is published in books, abstracts, news-
                                                                 letters and articles in major business, environ-
                                                                 mental and industry publications.
                                                                 I.S. Resources, Inc.
                                                                 4361 Route #8
                                                                 Allison Park, PA 15101
                                800-327-5895
HAZMAT SUPPLY GROUP/ISR is a one-stop
supplier of hazardous materials handling equip-
ment and training.  ISR uniquely  brings a wide
variety of products and services to the end user,
including a "total" Level A protection system by
TRELLEBORG/INTERSPIRO, specialized com-
munication equipment, program development, on-
site  equipment  trailers,  decon  trailers, and
advanced  HAZMAT training courses.
                                                                 In-Situ Inc.
                                                                 P.O. Box I
                                                                 Laramie, WY 82070-0920
                                                                                                                                 307-742-8213
In-Situ, Inc. began as a consulting firm providing
computer modeling, hydrologic evaluation, and
laboratory R & D services in in-situ mining and
related energy industries. In-Situ has broadened
its technological base to include the development
and sale of state-of-the-art hydrologic instrumen-
tation and software, to provide time sharing serv-
ices to clients through its Computer division.

Industrial & Environmental
Analysts, Inc. (IEA)
P.O. Box 12846
Research Triangle Park, NC
27709                           919-467-9919
Industrial & Environmental Analysts,  Inc. (IEA)
is a multidisciplinary certified analytical and con-
sulting laboratory working primarily with indus-
trial and governmental clients  in environmental
and chemical monitoring, testing, research and de-
velopment. lEA's nationally recognized capabili-
ties combine EPA Contract Laboratory Protocol
with State-of-the-Art Instrumentation in order to
provide efficient solution to analytical problems.
Inside Washington Publishers
1235 Jefferson Davis Highway, #1206
Arlington, VA 22202             703-892-8504
Inside Washington Publishers, a Washington, DC
based newsletter  publisher, offers up-to-the-minute
news coverage on federal and state environmen-
tal activity,  including legislation,  litigation and
regulation.   Inside EPA Weekly  Report,  the
Environmental Policy Alert, and the Superfund
Report as well as a document retrieval service and
conference group offering policy forums, comprise
the diverse information services provided by In-
side Washington Publishers.

Institute of Chemical Waste Management
1730 Rhode Island Avenue, NW Suite 1000
Washington, DC 20036           202-659-4613
ICWM is the Institute of Chemical Waste Manage-
ment. The Institute is a component of the National
Solid Wastes Management Association and was
formed in  the  late 1970's to  promote proper
management of hazardous waste. Members of
ICWM are those commercial firms engaged in all
aspects of hazardous waste management: trans-
portation, storage, treatment,  incineration and
disposal.

Intech Biolabs,  Inc.
158 Tices Lane  East
Brunswick, NJ  08816            201-257-1050
Intech Biolabs, Inc., is a privately owned indepen-
dent testing laboratory providing services for food,
pharmaceutical, cosmetic, environmental,  and
chemical industries. Intech Biolabs, Inc. has but
                                                                                                                     EXHIBITORS     543

-------
one guideline: Provide complete customer satis-
faction with accurate data analysis within  the
quickest possible turnaround time, at competitive
costs.

Integrated Data Technologies, Inc.
4813 Springfield Ave.
Philadelphia, PA 19143          215-726-6124
Integrated Data Technologies, Inc. (IDT) is a soft-
ware consulting and development company active
in the areas of environmental engineering and
hazardous waste  management.  IDT markets a
hazardous waste  manifest tracking system and
project management  software which specifically
addresses hazardous waste site cleanup. Based on
client needs, IDT will modify existing packages or
develop custom programs.

International Technology Corp
2790 Mosside Boulevard
Monroeville, PA 15146          412-243-3230
International  Technology  Corporation  is an
environmental management company providing a
range of services dedicated to the assessment, miti-
gation and decontamination of situations involving
hazardous substances. The company serves indus-
try and government with a project managemeni
capability which addresses environmental problem-
solving in five areas: risk control services, analy-
tical services, engineering services, decontamina-
tion and remedial services, and transportation,
treatment and disposal services.

International Waste Technologies
807 N. Waco, Suite 31
Wichita, KS 67203
                                316-683-8986
Chemical fixation products and Japanese in-situ
mixing drills.

Jacobs Engineering Group, Inc.
529 14th Street,  NW, Suite 1234
Washington. DC 20045           202-783 -1560
Jacobs Engineering Group, Inc., an international
environmental engineering and construction firm,
has extensive experience, particularly with respect
to the remediation of hazardous waste sites, and
provides engineering services to evaluate sites, and
characterizes the nature and extent of contamina-
tion, and develops remedial action plan alterna-
tives, including cleanup  design.

Junes T. Warring Sons, Inc.
4545 "S" Street
Capitol Heights, MD 20743       301-322-5400
All types  and  sizes of containers—new and
reconditioned—fiber, steel, plastic. Our hazardous
waste containers are DOT approved and range in
size from 5 to 110 gallons. We accept orders from
one to truck loads and  we ship anywhere. You
order  a  container—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. Also provid-
ed is empty drum removal with custom shredding
and crushing done on your site.
 Johnson Screens
 P.O. Box 64118
 St. Paul, MN 55164
612-636-3900
 Ground water  monitoring  well  completion
 materials, including screen, casing and sampling
 devices.
 Lancaster Laboratories, Inc.
 2425 New Holland Pike
 Lancaster, PA 17601
717-656-2301
 Lancaster Laboratories, Inc., is an independent
 analytical laboratory offering high quality tech-
 nical services with  personal attention to client
 needs. Our company employs over 200 people in
                its environmental and health science division. Our
                environmental services included the analysis of
                water, soil and hazardous waste for inorganic,
                organic, and traditional wet-chemistry parameters.
                Industrial hygiene  services,  including on-site
                sampling, are also offered. Instrumentation in-
                cludes: GC/MS, GC, purge and trap GC, HPLC,
                flame and furnace AA and ICP.
                Lars Lande Manufacturing
                11615 North Shore Road
                Whitmorc Lake, MI 48189
                               313.449-4116
                Manufacturer of rotary extractors and zero head
                space extractors  for laboratory  analysis of
                hazardous waste.
                Law Environmental, Inc.
                112 Town Park Drive
                Kennesaw, GA 30144
                               404-421-3390
                Law  Environmental,  Inc.  ...a  professional
                engineering and  earth science consulting firm.
                Services Include: Hazardous and Solid Waste
                Management,  Surface  Water  Hydrology  and
                Water Quality Protection, Ground-Water Hydrol-
                ogy and Resource Development, Land Treatment
                of Wastes, Ecological Assessments and Industrial
                Siting, Geophysical Exploration, Seismic Hazard
                Evaluation, Site  Remediation Management.

                Lawler, Matusky, & Skelly Engineers
                One Blue Hill  Plaza
                P.O. Box  1509
                Pearl  River, NY  10965            914-735-8300
                Lawler, Matusky  & Skelly Engineers is an environ-
                mental science and engineering  consulting firm
                founded  in  1965.  LMS  offers comprehensive
                services in waste management and groundwater
                resource planning including consultation and as-
                sistance in developing regulatory compliance pro-
                grams. LMS provides contamination assessment;
                preliminary site assessment;  field sampling and
                monitoring for wastes, soils, groundwater and sur-
                face waters; remediation design; environmental
                and risk assessment; modeling and statistical evalu-
                ation  of data; permit assistance; QA/QC com-
                pliance and  procedures; oversight services; and
                other  engineering/biological and  groundwater
                Layne-Western Company, Inc.
                1900 Johnson  Drive
                Mission Woods,  KS 66205
                                913-362-5440
                Layne-Western Company, Inc. brings technical
                knowledge and practical experience to the special-
                ized fields of investigative drilling, remedial action
                and environmental monitoring.  From  offices
                located coast-to-coast, we provide clients with a
                pool of talented professionals and a high commit-
                ment to professionalism,  safety and quality.
                Livingston Enterprises
                2855 Kifer Road, Suite 103
                Santa Clara, CA 95051
                                408-986-8866
Livingston Enterprises provides software programs
to assist companies in complying with State and
Federal Hazard Communication Standards and
Hazardous Substance  Inventory controls and
reporting requirements in accordance with SARA
  Title  III.
                Lopal Enterprises, Inc.
                1750  Bloomsbury Avenue
                Wanamassa, NJ 07712
                                201-922-6600
Manufacturer of K-20 brand products for the con-
trol and prevention of leaching and migration of
hazardous toxic substance. For applications in soil-
like paniculate matter and on various ccmentitious
surfaces. The easy-to-apply proprietary products
are recommended for use in the control of PCB's
and other chlorinated organic compounds, and
toxic  heavy metal;  lead, mercury,  chromium,
barium, arsenic, cadmium, etc.

MAC Corporation/Saturn Shredder Division

201 East Shady Grove Road
Grand Prairie, TX 75050         214-790-7800
MAC Corporation's Saturn Shredder Division
manufactures  low-speed,  high-torque,  rotary
shear-type shredders designed to produce required
size reduction application for light metals, ferrous
and non-ferrous,  plastic, wood, rubber, glass,
paper for recycling purposes and provide resource
recovery  or  waste  to  energy  assistance  for
hazardous, nuclear, municipal or industrial solid
waste. Innovative effective systems for  proper
waste reduction is Saturn's expertise.
                                                MAECORP Incorporated
                                                17450 S.  Halsted Street
                                                Homewood. IL 60430
                                312-709-0333
                                                MAECORP Incorporated is a full-service environ-
                                                mental management company committed to imple-
                                                menting the most cost-effective solutions to waste
                                                management by utilizing on-site treatment tech-
                                                nologies. MAECORP further has the capabilities
                                                to respond to emergency spills 24 hours a day. Cur-
                                                rently, MAECORP is performing several cleanups
                                                for the U.S. EPA as the prime contractor under
                                                the ERCS Region V Contract.
                                                MARCOR of DC, Inc.
                                                11557 Edmonston Road
                                                BeltsviUe, MD 20705
                                                                                                                                301-937-4858
                                                MARCOR is a firm specializing in asbestos abate-
                                                ment services. We have worked  hard to acquire
                                                the reputation as the LEADERS IN ASBESTOS
                                                ABATEMENT.  A  firm  with a proven  track
                                                record, MARCOR has the expertise, professional
                                                staff,  and  equipment to handle any asbestos
                                                problem.
                                                MFC EavironmeflUl
                                                8631 West Jefferson
                                                Detroit, Ml 48209
                               313-843-3350
Hazardous material cleanup contractor, ground-
water cleanups and surveys. High capacity, high
head, high viscosity pumping specialist. Removing
viscous materials from lagoons or tankage our
forte.

MSI Detoxification Incorporated
100 Erik Drive
Bozeman. MT 59715             406-58&4885
MSI Detoxification Incorporated (MDI) is a full-
service hazardous waste site detoxification  com-
pany. MDI's services extend from legal and tech-
nical problem definition, to integrated emergency
responses/remedial investigations, to feasibility
studies based on state-of-the-art analyses of site
parameters and alternative technologies, to total
site detoxification based on an unique biological
technology and complementary technologies which
will significantly lower detoxification costs. Also
call 201-864-2055.
MTA Remedial
(MTARRI)
1511 Washington Avenue
Golden, CO 80401
                                                                                                                                 303-279-4255
MTA Remedial Resources, Inc. is a remedial ac-
tion contractor  that  provides:  safe and  cost-
efficient execution  of remedial action projects
•  application of alternative technology available
today  • detoxification of contaminated material
on-site or in-situ to reduce the volumes requiring
handling  • reduction  of long-term  liabilities
•  recovery of useful or valuable minerals.
 544     EXHIBITORS

-------
Marcel Dekker, Inc.
270 Madison Avenue
New York, NY 10016
                                212-696-9000
Marcel Dekker, Inc., will be displaying new titles
including: Inhalation Toxicology (Salem), Eco-
nomic Methods for Multipoint Analysis and
Evaluation (Baasel), Materials Recovery From
Municiple Waste (Alter), and Reducing the Car-
cinogenic Risks in Industry  (Deisler). Discount
order forms are available at the booth.

Mary Ann Liebert, Inc. Publishers
1651 Third Avenue
New York, NY 10128            212-289-2300
Mary Ann Liebert, Inc. publishes journals, books
and news publications in the most exciting and new
areas of science and medicine. One of the jour-
nals, Hazardous Waste & Hazardous Materials,
edited by Norman Beecher, Sc.D., is the  official
journal of  the Hazardous  Materials Control
Research Institute.  This journal is  the  central
source of information for advancing technology,
providing economical and ecological methodology
for the regulation  and management of hazardous
waste and related hazardous material.

Mean Free Path Corporation
220 Pegasus Avenue
Northvale, NJ 07647             201-767-7300

Med-Tox Associates, Inc.
 1431 Warner Avenue, Suite  A
Tustin, CA 92680               714-259-0620
MED-TOX associates, multidisciplinary environ-
mental health service firm operating  from multi-
ple  offices  offering consultation services  for:
Toxicology Risk Assessment  • Hazardous Waste
Management  • Health  and  Safety Training
• Occupational  Medicine  Services  •  Medical
Standards and Physical Fitness  Testing • Legal
Support  Services  • Organic  and Inorganic Ana-
lytical Services • Optical and  Electron Microscopy
Services.
                Millipore Corporation
                80 Ashby Road
                Bedford, MA 01730
                                617-275-9200
 metaTRACE, Inc.
 13715 Rider Trail North
 Earth City, MO 63045
314-298-8566
 metaTrace, Inc. is an analytical laboratory estab-
 lished in 1986 in St. Louis, Missouri. metaTrace
 offers full service capabilities for organic, inorgan-
 ic and radiochemistry analyses of air, ground-
 water, surface water, wastewater, potable water,
 soil,  hazardous wastes and  biological  samples.
 Routine  analyses  include  organics/inorganics,
 toxics including dioxins/furans, mixed waste or
 co-contaminated waste, radiochemistry, TCLP,
 Appendix VIII and IX, explosives (military com-
 pounds), industrial hygiene, air quality (including
 odor characterization), methods development/vali-
 dation, and EPA priority pollutant, RCRA, and
 SARA analyses.

 Metealf & Eddy, Inc.
 P.O. Box 4043
 Woburn,  MA 01888              617-246-5200
 Engineers  and  planners:  Hazardous Waste
 Management • Program Management •  Contract
 Ops  • Privatization  • Sludge  Management
 •  Water and Wastewater Management • Roads
 /Bridges • Transportation • Facilities.
Midwest Water Resource
320 W. Santee Highway
Charlotte, MI 48813
517-543-8155
MWRI specializes in innovative solutions to en-
vironmental problems. MWRI's newest develop-
ment, VAPORTECH, is a proprietary .process to
inexpensively remove VOC's  from the vadose
zone.
                In addition to offering a broad range of products
                in the  membrane  filtration  area,  Millipore
                manufactures products in the following areas: en-
                vironmental  testing,  air monitoring,  microbio-
                logical testing, hazardous  waste testing, and
                sampling products.

                Mine Safety  Appliances Company
                P.O. Box 426
                Pittsburgh, PA 15230            412-967-3204
                Manufacturer of full line of personal protective
                equipment including  respiratory protection,
                clothing, and instrumentation for levels A through
                D.

                Modern Industrial Plastics
                3337 North Dixie Drive
                Dayton, OH 45414               513-276-6400
                Modern Industrial Plastics exhibits many products
                made of TEFLON  R  for Groundwater Moni-
                toring. These products include MIP casing and
                screen sold through stocking distributors, a vari-
                ety of high-capacity bladder pumps for purging
                and sampling, portable pump controller and clear
                bailers of TEFLON with new accessories.
                Monoflex, Inc.
                P.O. Box 4421
                Clearwater, FL 34618
                                800-257-5183
Monoflex, Inc. is a manufacturer of quality PVC
well screens and casing. In addition to well screens
and casings, we also manufacture bailer samplers,
centralizers, washdown valves and locking well
protectors. We will custom slot well screens to user
specifications.

Morris Industries Inc
777 Route 23
Pompton Plaines, NJ 07444      201-835-6600
Manufacturer of Morris Monitor well locking cap
which is designed to provide maximum security.
Made tamper-proof from  welded heavy  gauge
steel, it is available in above ground and flush
mount models. Both will accommodate your own
standard security padlock. Distributors of ben-
tonite products in  granular, powder  and  pellet
form.

NAPPI Trucking Corporation
P.O. Box 510
Matawan, NJ 07747              201-566-3000
Hazardous materials and  waste  transportation
services.

NERI/Petrex
309 Farmington Ave., Suite A100
Farmington, CT 06032           203-677-9666
The Petrex Technique is an innovative geochemi-
cal method for identifying  and mapping volatile
organic compounds from soils and groundwater
contamination.  The Technique utilizes  Petrex
monitors which are placed in the ground in a stra-
tegic pattern. After a  representative subsurface
sampling period, the monitors are removed and
analyzed by mass spectrometry.

NIOSH
4676 Columbia Parkway
Cincinnati, OH 45226            513-533-8323
Employees work safely at a hazardous waste site
if they  are informed  of the hazards involved,
receive necessary training, follow the proper  proce-
dures, use the required personal protective equip-
ment  and  remain  aware of the conditions  or
situations around them at all times. The NIOSH
exhibit includes recommendations for the control
of occupational hazards, including the NIOSH
Hazardous Waste Sites and Hazardous Substance
Emergencies-Worker Bulletin. Information will be
available for NIOSH Recommended Criteria, Cur-
rent Intelligence Bulletins, Technical Reports and
our data base.

NL Baroid/NL Petroleum Services Inc
P.O. Box 1675
Houston,  TX 77251              713-987-4850
NL Baroid's high quality bentonite products offer
cost-effective solutions  to soil permeability con-
trol.  Whether  soil  sealing,  slurry trenching,  or
other environmental applications, our technical
services will determine optimum treatment to meet
your project  requirements. For 60 years, NL
Baroid has devoted its resources in providing
leading edge Bentonite  technology.

NUS Corporation
Park West Two, Cliff Mine Road
Pittsburgh, PA 15275             412-788-1080
For 27 years NUS Corporation has provided the
environmental and  engineering expertise to solve
industry  and  government  waste management
problems with  cost-effective solutions. Our staff
of 1600 multidisciplinary professionals offer a full
range of services including: environmental assess-
ment, environmental engineering, remedial design
engineering, hydrogeologic and geologic services,
risk assessment, regulatory assistance, environmen-
tal health  and  safety and analytical services.

Nanco  Environmental Services, Inc.
Road 6 Robinson Lane
Wappengers Falls,  NY  12590      914-221-2485
Nanco Laboratories, a USEPA,  N.Y.D.E.C.,
N. J.D.E.P., and commonwealth of Virginia con-
tract laboratory provides complete environmen-
tal  analytical  services  nationwide.  NANCO
Express Data Service (EDS) includes independent
data audits, data interpretation service, electronic
data  delivery and guaranteed TAT.
National Draeger, Inc.
Post Office Box 120
Pittsburgh, PA 15230
                                                                                                                                412-787-8383
                                                                National Draeger, Inc., a U.S.  subsidiary of
                                                                Draegerwerk AG, located in Luebeck, West Ger-
                                                                many, has earned a worldwide reputation for being
                                                                the leader in manufacturing specialized equipment
                                                                and systems which enable, support and protect
                                                                human breathing safety. By introducing innova-
                                                                tive new products for gas detection and warning
                                                                systems, breathing protection, filter technology,
                                                                diving equipment, air supplied systems for avia-
                                                                tion and space technology as well as medical equip-
                                                                ment, Draeger helps to upgrade safety standards
                                                                for the entire safety  industry.

                                                                National Environmental Testing,  Inc
                                                                200 Grove  Road
                                                                Thorofare, NJ  08086            609-848-3939
                                                                A national network of environmental laboratories
                                                                offering a consistently high  standard  of service
                                                                across our  seven locations throughout the U.S.
                                                                Complete range of the  state-of-the-art analytical
                                                                instrumentation. We also offer field  sampling,
                                                                aquatic  toxicology  and  stack testing services.
                                                                Laboratories in  California, Illinois,  Indiana,
                                                                Texas, New Jersey.
National Lime Association
3601 North Fairfax Drive
Arlington, VA 22201
                                                                                                 703-243-5463
                                                                 The National Lime Association exhibit provides
                                                                 information on the use of lime for hazardous waste
                                                                 treatment. The principal use of lime is the neutrali-
                                                                                                                      EXHIBITORS     545

-------
zation of inorganic acidic waste and the precipi-
tation of heavy metals. Lime and  flyash for a
pozzolanic material which can be used to solidify
a hazardous sludge.
                 Orion Research
                 529 Main Street
                 Boston, MA 02129
                                617-242-3900
Northeastern Analytical Corp.
234 Route 70
Medford, NJ 08055
609-654-1441
Northeastern Analytical Corp. provides technical
support for hazardous waste projects. The Field
Services  Group  is experienced in all  facets  of
air/soil/waste sampling and tank testing. The
NAC Industrial Hygiene Group can manage the
health and safety program. The laboratory pro-
vides complete chemical analysis of site samples.

Nytest Environmental Inc.
60 Sea view Blvd.
P.O. Box 1518
Port Washington, NY 11050      516-625-5500
An  independent laboratory providing complete
analytical services for the screening and analysis
of hazardous waste. The laboratory is equipped
with sophisticated computerized GC/MS systems
including extensive automated allied instrumenta-
tion enabling it to perform analytical programs
thoroughly, competently and quickly.

 O.H. Materials Corp./Environmental
 Treatment & Technologies Corp.
 16406 U.S. Route 224, East
 Findlay, OH 45840              419-423-3526
 OHM and ETC are subsidiaries of Environmen-
 tal Treatment and Technologies Corp. (ETTC)
 offering an unmatched scope of on-site, environ-
 mental  services  nationwide:  groundwater/soil
 treatment,  facilities decontamination,  mobile
 incineration, data management, analytical services,
 surface impoundment restoration, UST manage-
 ment, waste site cleanup, emergency response and
 explosive handling.

ON-SITE INSTRUMENTS/
EnviroRENTAL
542  South Drexel Avenue
Columbus, OH 43209          1-800-551-2783
On-Site Instruments/EnviroRENTAL sells, rents,
leases and services a complete  line of industrial
hygiene,  laboratory, and environmental instru-
ments, equipment,  and supplies.  In  Ohio call
1-800-551-3102.

Occupational Hazards Magazine
1100 Superior Avenue
Cleveland, OH 44114
216-696-7000
Occupational  Hazards—edited for  officials
responsible for occupational safety/health and
plant protection. Covers legislative,  regulatory,
scientific, and  other developments affecting the
field, plus "how-to" articles. Continuing coverage
of  safety management,  industrial  hygiene,
occupational health, hazardous materials control,
incentives, training/education, insurance, fire pro-
tection, emergency response, security. Material
based on interviews with business, government
labor, academic and other experts.

Ogden Environmental  Services
Post Office Box 85178
San Diego, CA 92138            619-455-3045
Ogden Environmental  Services  will provide a
service for the disposal of toxic and hazardous
waste to  both private industry and government.
Major activities will include: Transportable com-
bustor for on-site remediation, fixed-site units for
disposal of continuous waste streams, fixed-site
regional incinerators for small generator waste dis-
posal. OES will build,  own and operate modular
facilities  that will be transportable to hazardous
waste sites for cleanup.
Orion manufactures pH  and ion-selective elec-
trodes, meters, and systems for fast, accurate, and
cost-effective  chemical measurement. Offering
portable, hand-held meters for field work and ad-
vanced systems for the Lab. Our versatile ORION
960 Autochemistry system performs a variety of
analytical techniques including ISE measurements
and potentiometric titrations.

P.E. LaMoreaux & Associates
P.O. Box 2310
Tuscaloosa, AL 35403           205-752-5543
P.E. LaMoreaux and  Associates, Inc. (PELA),
consulting hydrologists, geologists, engineers and
environmental scientists, offer hydrological, geo-
logical, environmental and hazardous waste con-
sultation services. Other services provided include
sampling,  laboratory  analysis, development  of
monitoring programs  and installation of  wells,
reclamation, government  permitting,  court tes-
timony,  and graphic  and communication
programs.

PRC Environmental Management, Inc.
303 E. Wacker Drive
3 Illinois Ctr.
Chicago, IL 60601              312-938-0300
PRC EMI provides environmental services to both
government  and  industry.  Headquartered  in
Chicago, IL,  PRC  EMI  maintains  offices  in
McLean, VA; Atlanta, GA; San Francisco, CA;
and Houston, TX. Specialties include remedial in-
vestigations/feasibility  studies,  endangermem
assessments, remedial design and implementation,
compliance audits,  permitting support,  waste
reduction audits, risk management support, envi-
ronmental and systems engineering, and program
management support.

Parker Systems
P.O. Box 1652
Norfolk, VA 23501              804-485-2952
Manufacturers and distributors of products for
control and clean up of oil and hazardous material
spills. Products list includes response and perma-
nent containment boom, light and heavy oil sor-
bents,  hazardous  material sorbenls,  skimmers,
washdown & transfer pumps, boom marker lights,
trailers, and hand tools.

Peroxidatlon Systems, Inc.
4400 E. Broadway, Suite 602
Tucson, AZ 85711              602-327-0277
Peroxidation Systems, Inc. provides equipment,
technical services and chemicals for the treatment
of contaminated water and concentrated organic
wastes by means of chemical oxidation destruc-
tion using  proprietary UV  hydrogen peroxide
process.

Phoiovac International Inc.
741 Park Avenue
Huntington, NY 11743           516-351-5809
AUTOTIP designed  to  continuously monitor
chemical pollutants in air from 0.1 to 2,000 ppm.
an internal pump can draw samples  in from  up
to 1,000 ft. Calibration is automatic there are two
alarm levels with option of RS232 communications
and multipoint sequential sampling. Also displayed
 10S series portable G.C.s and TIP I, II, and HI.

Pollution Abatement Consultant!)
800 Orange Street
P.O.  Box 1039
Millville, NJ  08332               609-825-5575
Manufactures sampling equipment  for drums,
tanks, surface and  groundwatcrs, sludges and
solids. Also provides  safety-coated sample con-
tainers, sample shipment systems, environmental
laboratory kits, apparatus and accessories as well
as personal protective equipment.
Poly-America Inc.
2000 W.  Marshall
Grand Prairie, TX 75051         800-527-3322
Poly-Flex is a polyethylene geomembrane liner
(20-100 mils thick) which provides a cost-effective
method of lining hazardous waste disposal facili-
ties and preventing groundwater pollution. Poly-
Flex is manufactured by Poly-America, one of the
most  modem extrusion facilities in the U.S.,
producing 150 million pounds of polyethylene per
year. Poly-America's state-of-the-art quality con-
trol laboratory assures the finest quality polyethy-
lene liner available.
                                                                Precision Laboratories
                                                                Post Office Box 915
                                                                Garden Grove, CA 92642
                                714-891-7832
                                                                Precision Laboratories specializes in physical and
                                                                compatibility  testing of  geosynthetics.  Fast,
                                                                accurate results are provided on geomembranes,
                                                                geotextiles, geogrids and thermptastic piping. Spe-
                                                                cial services include: liner/waste compatibility
                                                                studies such as EPA Method 9090, sameday test-
                                                                ing of seams, and quick turn-around time on qual-
                                                                ity assurance testing.
                                                                Preferred  Reduction Services, Inc.
                                                                930 F. Calle Negocio
                                                                San Clemente, CA 92672         714-498-8090
                                                                PRS,  Inc. specializes in the physical and chemi-
                                                                cal treatment of hazardous waste and septage from
                                                                domestic, food, and industrial areas, as well as site
                                                                cleanup activity. PRS provides belt presses, phys-
                                                                ical chemical treatment units, and incinerators for
                                                                almost every hazardous application.
                                                                Pressure Filtration Specialists
                                                                30 Mason Street
                                                                P.O. Box  686
                                                                Torrington. CT 06790            203-489-1221
                                                                PFS trailer-mounted  filter presses  can  reduce
                                                                generator liquid sludge volume by producing filter
                                                                cake of 35-60 percent solids by weight. They can
                                                                handle large as well as small jobs, and can pro-
                                                                vide a 24  hour/day operation processing up  to
                                                                250,000 gal/wk of generator waste depending upon
                                                                composition of sludge.

                                                                Princeton Testing Laboratory, Inc.
                                                                Post Office Box 3108
                                                                Princeton, NJ 08540
                                609-452-9050
Environmental analysis and industrial hygiene.
Toxic waste/soil, RCRA,  NJ ECRA, industrial
wastewater NPDES, ground water, OSHA work-
place surveys, asbestos monitoring and evaluation,
complete NIOSH lab methodology, stack testing,
asbestos abatement training courses, Right-To-
Know  compliance, microbiology. A1HA accre-
dited, NJ DEP Certified, NY DOH approved, PA
DER approved.
QED Environmental Systems
Post Office Box 3726
Ann Arbor, MI 48106           313-995-2547
Well Wizard*  dedicated groundwater sampling
systems, Sample Pro-portable samplers and sup-
plies, and Pulse Pump*  pneumatic pumping sys-
tems  for  leachate  and  contaminated  water
pumping are all featured in the Q.E.D. Environ-
mental Systems, Inc. line. Well Wizard*  dedi-
cated  bladder  systems and  purging aids are
designed  to meet  the latest  EPA  sampling
standards.
Quallmelrics, Inc.
P.O. Box 230
Princeton, NJ 08542             609-924-4470
Qualimetrics manufactures a complete  line of
meteorological monitoring sensors and systems.
546     EXHIBITORS

-------
These range from stand-alone, battery-powered
portable weather stations to Automatic Weather
Observing Systems (AWOS) stations for use at air-
ports around the world.

R.E. Wright Associates
3240 Schoolhouse Road
Middletown, PA 17057          717-944-5501
R.E. Wright Associates, Inc. (REWAI) is an earth
resource consulting services firm specializing in
providing technical consulting services in the areas
of hazardous waste management, hydrogeology,
environmental engineering, solid waste disposal,
soil science, engineering geology, wastewater dis-
posal,  and mining geology.

REMCOR, Inc.
701 Alpha Drive, Box 38310
Pittsburgh, PA 15238            412-963-1106
REMCOR is a turnkey hazardous waste contrac-
tor offering a full line of services including Audits,
Sampling, Groundwater Studies, Analysis, Design,
Equipment Decontamination, Soil Cleanup, and
Landfill and Lagoon Closure. Remcor performs
studies of waste disposal options and then imple-
ments  its recommendations by performing the
remediation.
REWAI Environmental Restoration
Systems
3240 Schoolhouse Road
Middletown, PA 17057
800-238-3320
REWAI  Environmental  Restoration  Systems
designs and manufactures groundwater cleanup
equipment,  including the Auto-Skimmer,  Air
Stripping Towers, Water Table Depression Pumps,
Liquid Interface Samplers, and Syphonid.  The
Auto-Skimmer automatically recovers subsurface
spills of floating hydrocarbons from both large and
small diameter wells.
RMT, Inc.
1406 East Washington, Suite 124
Madison, WI 53703              608-255-2134
RMT, Inc. specializes in solid and hazardous waste
management, remedial  investigation/feasibility
studies, groundwater quality, underground tank
management, wastewater treatment, air pollution
control,  industrial hygiene  engineering,  and
analytical services. Offices are located in Madison,
Wisconsin; Grand Ledge, Michigan; Greenville,
South Carolina;  Washington, DC;  and Santa
Monica,  California.

Radian Corporation
5103 W.  Beloit Road
Milwaukee, WI 53214            414-643-2668
Providing technical and engineering solutions to
environmental problems, such as:  wastewater
treatment, control of toxic liquids and air pollu-
tants, water reuse, groundwater remediation, land-
fill leachate treatment,  and  EPA  Superfund
projects;  including complete, certified environ-
mental laboratory services.

Recra Environmental Inc.
10 Hazelwood Drive, Suite 106
Amherst, NY 14075              716-691-2600
Recra Environmental, Inc. is an environmental
consulting and testing firm specializing in chemi-
cal waste analysis, prevention and control. Our
specialized  services include data  analysis,  data
management services, chemical testing, real estate
transfer assessments, waste minimization programs
geared towards source  reduction and treatment
feasibility studies.

Resource Analysts, Inc.
P.O. Box 778
Hampton,  NH 03842            603-926-7777
Resource Analysts, Inc. and its affiliates provide
comprehensive environmental testing service to in-
dustrial and commercial clients and to all levels
of government. Specialties include organic chemis-
try using IR, GC, and GC/MS analytical methods;
inorganic and heavy metal chemistry; freshwater
and marine aquatic toxicology; and field sampling.
The  laboratories  occupy a 20,000 square foot
facility with a staff of 65 professionals and par-
ticipate in the EPA Contract Laboratory Program.

Response Rentals
1460 Ridge Road East
Rochester, NY  14621            716-266-3910
Response Rentals brings you: instruments to sup-
plement your own; emergency equipment when
yours is down; or special items for that one-time
only use. Offers environmental consultants a low-
cost way to expand their services. Product line in-
cludes organic vapor analyzers, particulate sam-
plers, mercury vapor monitors, and more.

Riedel Environmental Services, Inc.
P.O. Box  5007
Portland, OR 97208-3320        503-286-4656
RES  provides  diverse  environmental  services
including  environmental consulting and engi-
neering, groundwater, remedial action and studies,
hazardous waste management, emergency response
for oil  or  hazardous  waste spills  or  releases.
Engineers, designs, installs and operates petrole-
um hydrocarbon recovery systems. Provides serv-
ices for industry and government from  regional
offices in St.  Louis, MO; Portland, OR; Richmond
and Los Angeles, CA.
                 Rollins Environmental Services, Inc
                 One Rollins Plaza, P.O. Box 2349
                 Wilmington, DE 19899          302-479-3164
                 The nation's first and most experienced hazardous
                 waste management company, Rollins continues to
                 lead in the development and application of state-
                 of-the-art technologies and services. From consul-
                 tation to destruction, Rollins offers  a complete
                 spectrum of hazardous waste management options
                 at regional facilities located in industrialized areas
                 of the country.
                                                                 SAIC
                                                                 8400 Westpark Drive
                                                                 McLean, VA 22102
                                703-821-4749
                 Roy F. Weston
                 Weston Way
                 West Chester,  PA 19380
                                215-692-3030
                 Managers  of major environmental projects in-
                 cluding facilities design, construction, operations,
                 remedial actions, permitting, closures, investiga-
                 tions and analytical chemical analyses.


                 Ruska Laboratories, Inc.
                 P.O. Box 630009
                 Houston, TX 77236-0009         713-975-0547
                 Manufacture all fused quartz analyzers for rapid
                 field screening of priority toxic substances in soil
                 of sedimentary sludge. Potential for small mobile
                 laboratory to identify and quantify the on-site
                 presence of chemical substances on a screening or
                 field support basis. Techniques involve integrated
                 thermal extraction, capillary  GC and mass spec-
                 trometry.

                 S & ME
                 P.O. Box  1308
                 Gary, NC  27511                 919-481-0397
                 Soil & Material Engineers provides a comprehen-
                 sive range of environmental and geotechnical
                 engineering  services to  help you  solve  your
                 hazardous substance problems. We will interpret
                 your responsibilities under the regulations, inves-
                 tigate your hazardous substance  problems, and
                 provide you with a cost-effective solution. We are
                 ready to serve you-call us.
                                                                 SAIC is a full service engineering and scientific
                                                                 consulting  firm providing waste  management
                                                                 assistance to Government and commercial clients
                                                                 in RCRA compliance, waste site investigations and
                                                                 clean-up,  technology development,  and waste
                                                                 management assessments. SAIC engineers and
                                                                 scientists have hands-on experience with the most
                                                                 recent  regulatory requirements and  technology
                                                                 developments.
                                                                 SCS Engineers
                                                                 11260 Roger Bacon Drive
                                                                 Reston, VA 22090
                                703-471-6150
Solid and Hazardous Waste Consulting Services
since 1970. Specialists in control and treatment of
subsurface gases; a subsidiary of the firm offers
construction services in this area. Remedial Inves-
tigations, Feasibility Studies, Design, Construction
Management. Analytical laboratory, underground
tank testing, construction. Real Estate Contami-
nation Assessments.  Hazardous  Waste Facility
Permitting.

SMC Martin, Inc.
900 W. Valley Forge Rd.
P.O. Box 859
Valley Forge, PA 19482          215-265-2700
SMC Martin Inc. provides engineering and geo-
technical services to industry and government in-
cluding remedial investigations  and feasibility
studies for hazardous and  solid  waste disposal
sites; wastewater treatment; environmental assess-
ment, legal advisory services; permit application
assistance; environmental monitoring; construc-
tion  management; closure plan design; and en-
gineering/financial analyses.
                                                 Sentex Sensing Technology, Inc.
                                                 553 Broad Avenue
                                                 Ridgefield, NJ 07657
                                201-945-3694
Portable computer controlled gas chromatographs
for on-site toxic gas detection. Introduction of a
"PC" operated portable gas chromatograph in-
corporating a lap-type personal computer.

Serrot Corporation
7575 Reynolds Circle
Huntington Beach, CA 92647     714-848-0227
Serrot Corporation is an installer of High Den-
sity Polyethylene Liners in a wide range of appli-
cations from Waste Disposal to Mining. Serrot also
manufactures  and  installs  special PVC Tank
Liners.

Sevenson Containment Corporation
2749 Lockport Road
Niagara Falls, NY 14302          716-284-0431
Sevenson Containment Corporation provides turn-
key services to government and industry in the
areas of hazardous waste management and waste
site cleanup. Sevenson's full service capabilities
include: site restoration; secure landfill construc-
tion; slurry wall & trench construction;  sludge
solidification & fixation; waste recovery and treat-
ment; drum removal and waste excavation,  and
transportation and disposal.

Shredding Systems Inc.
P.O. Box 869
Wilsonville, OR 97070           503-682-3633
Shredding Systems, Inc., manufactures a full line
of rotary shear shredders for compaction and in-
cineration systems. Models are available in both
electric and hydraulic drives in sizes up to 400 HP
and 1-0" x 52" infeed. The low speed, high torque
shredders are used in various applications including
                                                                                                                      EXHIBITORS     547

-------
the shredding of 55-gallon metal drums filled with
hazardous/nuclear waste.
Silicate Technology Corporation
14455 North Hayden Road, Suite 218
Scottsdale, AZ 85260            602-948-1300
Reagents used in the cleanup of hazardous waste;
they detoxify and render hazardous waste non-
leachable. Used in a variety off applications, in-
cluding on-line industrial treatment, land reclama-
tion and organic selective spill control, the reagents
cost a fraction of existing methods.
Silidur North America Co.
P.O. Box 1043
Elyria, OH  44036                216-277-0981
An imbankment stabilization process for stable
and unstable soil conditions. It addresses all types
of slips and channel work very cost-effectively. It
functions as a flexible concrete retaining wall that
turns into a living wall.
Skolnlk Industries, Inc.
4601 W. 48th Street
Chicago, IL 60632wป896         312-735-0700
Manufacturer of hazardous material containers
from 8 to 96  U.S. gallon sizes. Featuring "Big
Mouth" the 85 Gallon Salvage Overpack Drum,
and "Quad Pak" the Assortment of 8, 20, 55, 85
gallon units that fit inside of one another. Con-
tainer accessories and  components available as
well.
                 TAMS Engineering
                 655 Third Avenue
                 New York. NY 10017
                                212-296-4371
Solinst, Canada Ltd.
2440 Industrial Street
Burlington, Ontario, CN
L7P-1A5
416-335-5611
Manufacturers  of  high quality groundwater
monitoring instrumentation, including: the Water-
loo Multilevel System for monitoring multiple, dis-
crete zones, in a single bedrock borehole; Solinst
Water Level Meters for repeatable accuracy; and
the new Solinst Well Casing Depth Indicators. Also
new: The Saturated Sand Sampler retains up to
97%, including pore fluids, of 5  ft-long core of
samples.
Steam Kit Corporation
P.O. Box 1686
737 East Main St.
Salisbury, MD 21801             301-749-9318
Hazardous waste management & transportation;
Waste oil service;  Emergency spill response;
Manufacturer of tank testing equipment; Under-
ground storage tank testing, cleaning, abandon-
ment; Cathodic protection retrofit; Groundwater
remediation; Monitoring well installation; EPA-
DOT  hazardous waste pin fed label  service.
Sybron Chemkซb, Inc.
Post Office Box 66
Birmingham, NJ 08011           609-893-1100
Leaders in the application of Augmented Biorecla-
mation (ABR) for the treatment of contaminated
soil and  groundwater. Capabilities  include bi-
osystems engineering services and supply of selec-
tively  adapted   organisms   for   specific
contaminants. Technology useful for cleanup of
chemicals from leaking storage tanks,  pipeline
spills, train derailments, etc. Advantages are ulti-
mate disposal technology and low cost.
 Syprotec Inc.
 88 Hymus Boulevard
 Pointe Claire, Quebec,  CN
 H9R-1E4                       514-694-3637
 The Qwik-Skrene soil analytical testing system for
 PCB oil spills uses an extration  agent, efficient
 mechanical shaking, and nitration to prepare ex-
 tract  for PCB detection. Compatible  with the KS-
 ATS for oil, a color change indicates the presence
 of PCB. Pen samples can be extracted, and up to
 35 extracts tested for PCB simultaneously.
TAMS, a leading international engineering firm,
offers  comprehensive  services  in  solid  and
hazardous waste management. Capabilities include
Rl/FS; QA/QC; Health/safety; Risk Assessment;
Community Relations;  Remedial Design; Con-
struction Supervision; Site Closure; waste Geotech-
nics;  Chemical/Process  Design;  Watershed
Management;  Hydrogeology/Mathematical
Modeling. TAMS provides services to clients in
government, military and private sectors through
offices in major cities.

TECHLAW, Inc.
12030 Sunrise Valley Dr.
Reston, VA 22091               703-476-1100
%  Information  Management  and  Analysis
• Liligaiton  Support  •  Legal and Technical
Research •  Environmental  Audits TECHLAW
specializes   in  hazardous  waste  information
management including PRP searches, administra-
tive record  management, waste transaction data-
bases (NBAR), document inventory databases, 104
(e)  response management, litigation information
management, evidence audits and environmental
audits. TECHLAW combines lawyers, scientists,
and engineers to bridge the issues of law and tech-
nology.

THERMO  Automation, Inc.
686 South Taylor Avenue
P.O. Box 925
Louisville, CO 80027-0925        303-665-9000
THERMO  is a manufacturer of portable seam
welding equipment for thermo-plastic liner appli-
cations. ROVER is an  automatic self-propelled
seam welder specifically designed for rough ter-
rain applications including hazardous waste con-
tainment liners, grain storage, gas tank  storage,
landfill, reservoir, pit liners. TightSpot is a hand-
held ROVER.

Target Environment*! Services, Inc.
8940-A Route 108
Columbia,  MD 21045-2122        301-992-6622
Target performs soil gas surveys for the detection
and delineation of subsurface contamination by
volatile organic compounds. This low-cost, quan-
titative screening and monitoring technique b ideal
for environmental assessments of gas stations, in-
dustrial sites, landfills, town gas sites,  military
bases,  and  properties  undergoing  ownership
transfers.

Taylor Technologies
31  Loveton  Circle
Sparks, MD 21152               301-472-4340
We offer water test kits, reagents and apparatus
for all types of analysis, particularly waste and
potable water.

Technoi, Inc.
3333 N.W. 21st Street
Miami, FL 33142                305-634-4507
Technos offers state-of-the-art consulting in the
applied earth sciences. Our specialties include
hydrogeologic assessment of contaminant flow at
complex sites including those with: fracture flow,
sinkhole and subsidence problems, and sand lenses
in glacial till. We have been involved with geotech-
nical problems since 1971  and hazardous waste
sites since 1977. This long experience has given us
the opportunity to observe a wide variety of com-
plex site conditions and devise unique approaches
to evaluate those  conditions. Our site investiga-
tion work is supported by the finest, in-house sur-
face and downhole geophysical capabilities in the
U.S.
Telra Tech, Inc.
(A Honeywell Subsidiary)
630 North Rosemead Boulevard
Pasadena. CA 91107             818-449-6400
Tetra Tech provides water resource and hazardous
material management services for industrial, in-
stitutional and governmental clients throughout the
U.S. The firm's hazardous waste management
services include: Materials management, waste
characterization, emergency response, risk assess-
ment, feasibility studies, geohydrologk investiga-
tions, SPCC/contingency plans, health and safety
training.
Texas Research Institute. Inc.
9063 Bee Caves Road
Austin. TX 78733                512-263-2101
A scientific  research and development company
that offers  1- Laboratory services for evaluating
geosynthetic materials for use in landfills; 2- Site
safety and OSHA compliance by A!HA accredited
lab;  3- Site specific treatability  studies  for bi-
oremediation and 4- Specialized RAD for innova-
tive solution to hazardous waste problems.

Thermo Analytical tec./Norcal
2030 Wright Avenue
P.O. Box 4040
Richmond. CA 94084X1040       415-235-2633
Thermo Analytical, IDC.  TMA is a national organi-
zation of environmental and  materials  testing
laboratories with offices and facilities coast to
coast. TMA provides a full range of analytical and
field services to the A & E profession, the indus-
trial sector and government agencies. Specialties
include priority pollutants, radionuclides and as-
bestos determinations; health physics,  industrial
hygiene surveys, and source testing by trained field
teams; product testing and failure analysis using
latest instrumentation.

Tigs Corporation
Post Office Box 11661
Pittsburgh. PA 15228            412-563-4300
Manufacturers of Activated Carbon Modular Ad-
sorption Systems for the removal of organks from
water and air. With over 30 years of experience
in carbon technology and systems. TIGG Corpo-
ration provides not only equipment but the tech-
nical expertise to assist the client in projecting
results and economics.

Toxk Treatments (USA) Inc.
901 Mariner's Island Boulevard #315
San Mateo, CA 94404            415-572-2994
The DETOXIFIER is a mobile treatment unit used
in the in-situ remediation of contaminated soils
and waste deposits. It can treat a wide range of
chemical contaminants, including volatile  and
metals. The Detoxjfier can deliver and mix treat-
ment processes and reagents (oxdidents, fixatives,
steam, bioregents) into the contaminated soil.

Treatment Technologies
Technology Center, 2801 Long Rd.
Grand Island, NY 14072         716-773-8660
TreaTek is an environmental remediation joint
venture between Occidental Chemical Corporation
and Biotal Ltd. of Cardiff, U.K. Based on several
years of demonstrated success on Europe and U.S.
by each venture partner, this formal combination
is chartered to implement on-site conventional and
microbial treatment technologies for toxic teachate
and  contaminated  soil.  TreaTek's  technology
package has been successfully proven at complex
superfund  sites, coal-tar sites,  and petroleum
refineries. Operating either as a subcontractor,
prime contractor, or the total site manager, our
team  of scientists and engineers provides  site-
specific, economical, safe, and innovative solutions
to hazardous and toxic waste problems.
 548     EXHIBITORS

-------
Tri State Motor Transit Co.
3200 Pulaski Highway
Northeast, MD 21901            301-287-2520
Tri-State Motor Transit Co. is a specialized motor
carrier, offering efficient, economical service to
shippers of packaged hazardous waste. A network
of strategically located terminals supports our fleet,
which operates over ninety million miles each year,
nationwide.  Our Pacific State Services Division
dedicates specially trained drivers and specialized
equipment to the needs of volume shippers.

Triangle Laboratories, Inc.
P.O. Box 13485
Research Triangle Park, NC
27709                           919-544-5729
Triangle Labs is a full service GC/MS lab. Analysis
of water, soil, sludge, hazardous waste, tissue, bio-
logical samples, and air pollution samples for pri-
ority  pollutants,  dioxins/furans,  and  other
organics a specialty. High resolution MS, EI/CI
MS, FAB/MS and LC/MC are also available.

U.S. Army Corps of Engineers
P.O. Box 103
Downtown Station
Omaha, NE 68101-0103          402-221-7317
The U.S. Army Corps of Engineers and the U.S.
EPA have joined forces to  clean up Federal lead
hazardous waste sites under the Superfund pro-
gram. The booth will be manned by Corps per-
sonnel to  assist  architect-engineer  firms and
construction^ contractors take advantage of work
available to them through the Corps of Engineers.
 U.S. Bureau of Mines
 2401 E Street, NW, M.S.  1041
 Washington, DC 20241
202-634-1224
 The Bureau of Mines  is a federal government
 agency  charged  with  conducting research and
 gathering data pertinent to the minerals industry.
 Some of its research is oriented toward making
 mining more compatible with the environment and
 controlling minerals wastes. The Bureau's exhibit
 at SUPERFUND '87 will highlight recent research
 designed to treat contaminated water and control
 metals migration. In addition, the exhibit will fo-
 cus on technology developed by the Bureau to treat
 metals contaminated at battery breaker sites.


 U.S. Environmental Protection
 Agency
 401 M Street,  SW  WH 548A,  Rm 3410
 Washington, DC 20460          202-382-5100
 The Superfund law provides the  authority to
 respond to problems at uncontrolled hazardous
 waste sites in  emergency situations and at sites
 where long-term permanent remedies are required.
 The Emergency Response Team  responds  to
 releases which pose an immediate threat  to the
 public health and environment and provides R&D
 assistance in providing long-term remedies. The
 long-term remedies are performed by states, the
 U.S. EPA via the Army Corps of Engineers, or
 responsible parties.
U.S. Envirosearch
445 Union Boulevard, Suite 225
Lakewood, CO 80228
303-980-6600
A nation-wide recruiting firm based in Denver,
Colorado,  specializing  in  the  recruitment  of
hazardous waste, environmental and incineration
personnel.  U.S. Envirosearch represents client
companies in the areas of: Hazardous waste dis-
posal, site remediation, environmental engineering,
analytical labs, solvent recycling, PCB disposal,
industrial cleaning,  and generators.
                U.S. Geological Survey
                790 National Center
                Reston, VA 22092
                                703-648-4377
                Panels depicting research and products of the U.S.
                Goelogical Survey dealing with earth sciences.
                USPCI, Inc.
                2000 Classen Center, Suite 400S
                Oklahoma City, OK 73106
                                405-528-8371
                Full  service  hazardous  waste  management-
                Disposal, transportation, remedial cleaan-up, PCB
                destruction, EPA approved analytical laboratory,
                incineration available in 1989.
                United States Testing Company Inc.
                1415 Park Avenue
                Hoboken, NJ 07030
                                201-792-2400
                The Chemical Services Division, (EPA CLP) and
                certified by NJ D.E.P., NY, DOH, and Chemi-
                cal Waste Management utilizes 21 laboratories and
                83 major pieces of analytical equipment housed
                in 20,000 square feet, located at Corporate Head-
                quarters in  Hoboken,,  NJ. Services  Include
                Chemistry, Biology,  Asbestos  and  Nuclear
                Chemistry.    Contact   Jack   McLachlan
                201-792-2400.
                VFL Technology Corporation
                42 Lloyd Avenue
                Malvern, PA 19355
                                215-296-2233
VFL Technology offers solidification and stabili-
zation services for organic and non-organic liquids,
sludges and soils. In situ stabilization provides per-
manent chemical encapsulation at a fraction of off-
site disposal costs. Our field construction groups
have developed and implemented cost-effective on-
site closures for impoundments ranging from 2-200
thousand cubic yards.


VNR Information Services
115 Fifth Avenue
New York, NY 10003            212-254-3232
CHEMTOX™ DATABASE-A  cost-effective,
easy-to-use, and portable means to access data on
over 3,600 hazardous chemicals. Books, journals
and newsletters on hazardous chemicals and in-
dustrial toxicology.


Versar
9200 Rumsey Road
Columbia,  MD 21045            301-964-9200
Environmental compliance management system.


WAPORA, Inc./Kemron Environmental
1555 Wilson Boulevard, Suite 700
Rosslyn, VA 22209              703-524-1171
If you haven't "heard it through the grapevine",
WAPORA offers a strong corporate commitment
to developing the most cost-effective solutions to
waste management problems. Catch our act at Ex-
hibit #1313. Whether we are providing ground-
water assessments, RCRA support programs,  or
cleanup management activities, we  always produce
satisfied clients. We provide turnkey environmen-
tal services from initial investigation, laboratory
analyses (through our affiliate, KEMRON) to final
cleanup and disposal management.
                 Wadsworth/Alert Laboratories, Inc.
                 1600 Fourth Street, SE
                 Canton, OH 44701
                                216-454-5809
Waste Conversion, Inc.
2951 C Advance Lane
Colmar, PA 18915               215-822-2676
Waste Conversion, Inc. offers services including
cleanup, oil reclamation, transportation, analysis,
treatment  and disposal of hazardous  and non-
hazardous materials. Support services include a
complete transportation fleet comprised of cacuum
tankers, box and dump trailers permitted to trans-
port hazardous and  non-hazardous  materials
throughout the eastern half of the country.
                                                Watersaver Company, Inc.
                                                P.O.Box 16465
                                                Denver, CO 80216
                                                                                                 303-289-1818
                 Analytical Laboratory, Mobile Labs, Regional
                 Locations. CLP Approved.
Watersaver custom fabricates geomembrane sys-
tems for a wide variety of applications, including
hazardous waste. PVC, CPE, HYPALON, XR-5
in thickness from 30 to 60 mils. Factory fabricated
accessories complete the Watersaver Lining Sys-
tem. Meet all current federal and state regulations
by relying on Watersaver. Continuous service for
over a quarter century.

Wehran Engineering
666 E. Main Street
P.O. Box 2006
Middletown, NY 10940           914-343-0660
Wehran,  an environmental consulting firm of en-
gineers and scientists, is recognized as one of the
top firms nationally  in  the field of solid and
hazardous waste management. Located in Midd-
letown, New York, with branch offices in western
New  York, New  Hampshire,  Massachusetts,
Indiana,  Vermont and Buenos Aires, Argentina,
Wehran employs 150 professionals who provide
a multi-disciplinary approach to  developing in-
novative  and cost-effective solutions to waste
management problems.

Westinghouse Environmental
Technology Division
Post Office Box 286
Madison, PA 15663              412-722-5116
Westinghouse provides solutions that help to find
a balance between  beneficial uses of industrial
processes and potentially negative impacts on our
environment. We provide proven,  advanced ther-
mal destruction technologies for  industrial and
hazardous waste management. Our comprehensive
services include project management, system de-
velopment, site management, and regulation per-
mitting.

Wilson Laboratories
525 N. 8th Street, P.O.  Box 1884
Salina, KS 67401                 913-825-7186
Full service analytical laboratory  specializing in en-
vironmental monitoring  and  the  analyses  of
hazardous  waste  samples.  Expertise includes
GC/MS, GC, HPLC, and Industrial Hygiene.

Woodward-Clyde Consultants
201 Willowbrook Boulevard
Wayne, NJ 07470                201-785-0700
Woodward-Clyde Consultants  is  a professional
services firm with over 30 years of experience in
geotechnical  engineering,  environmental,  and
social sciences. In hazardous waste, we offer total
management  solutions,  from  evaluation,  per-
mitting and initial investigation through design,
construction, and remedial action. Our scientists
and engineers represent all disciplines necessary to
provide complete services - the earth, physical, and
natural sciences as well as environmental, chemi-
cal and geotechnical engineering.  With offices in
30 cities, we are staffed and positioned to  offer
nationwide management programs that are both
comprehensive and responsive.
                                                                                                                     EXHIBITORS     549

-------
 YSI Incorporated                                ductivity in groundwater and surface water. It    pH and temperature in a flow-through chamber
 1725 Brannum Lane                              measures temperature from - 5 to + 50 C, water    that can be connected to monitoring weU pump-
 Yellow Springs, OH 45387       513-767-7241      level to 150 feet, conductivity or temperature-    ing syซtems.
                                                compensated conductivity to 20 millimhos/cm.
 YSI makes instruments for measuring ground and      The new YSI Model 3500 Water Quality Monitor    York Laboratories
 surface water quality. The YSI Model 3000 T-L-      measures conductivity, temperature-compensated    200 Monroe Turnpike
 C Meter measures temperature,  level and con-      conductivity, eh, pH, temperature-compensated    Monroe, CT 06468              203-261-4458
550     EXHIBITORS

-------
                                                  AUTHOR  INDEX
                       This  Author Index contains authors for  the years  1980 to 1986 only.
 Absalon, J. R., 80-53
 Accardi, J., 55-48
 Adamowski, S. J., 55-346
 Adams, R. B., 54-326
 Adams, W. M ., 55-108
 Adams, W. R., Jr., 52-377, 55-352
 Adaska, W. S., 54-126
 Adkins, L. C., 50-233
 Aguwa, A. A., 55-220
 Ahlert, R.  C., 52-203; 55-217; 54-393
 Ahnell, C.  P., Jr., 50-233
 Ainsworth, J. B., 55-185
 Albrecht, 0. W., 57-248, 393
 Aldis, H., 55-43
 Aldous, K., 50-212
 Alexander, W. J., 52-107
 Allcott, G. A., 57-263
 Allen, H. L., 57-110
 Alvi, M. S., 54-489
 Ammann, P., 54-330
 Ammon, D., 54-62, 498
 Amos, C. K., Jr., 54-525
 Amster, M. B., 55-98
 Anastos, G. J., 5(5-93, 322
 Anderson, A. W., 54-511
 Anderson,  D. C., 57-223; 55-154;
  54-131, 185; 55-80
 Anderson,  E. L., 55-193
 Anderson,  J. K., 54-363
 Andrews, J. S., Jr., 55-78
 Apgar, M., 54-176
 Appier, D. A., 52-363
 Arland, F.  J., 55-175
 Arlotta, S.  V., Jr., 55-191
 Arnold, D. F., 54-45
 Arthur, J., 54-59
 Asoian, M. J., 5(5-152
 Assink, J. W., 52-442; 54-576
 Astle, A. D., 52-326
 Atimtay, A., 55-464
 Atwell, J. S., 55-352
 Ayres, J. E., 57-359
 Ayubcha, A., 54-1

 Badalamenti.'S., 55-202, 358; 54-489
 Baer, W. L., 54-6
 Bagby, J. R., Jr., 55-78
 Bailey, P. E., 52-464
 Bailey, T. E., 52-428
 Bailey, W.  A., 55-449
 Balfour, W. D.,  52-334; 54-77
 Ballif,  J. D., 52-414
 Barbara, M. A., 55-237; 55-310
Barboza, M. J., 55-152
Bareis, D. L., 55-280
Barker, L.  J., 52-183
Barkley, N. P., 52-146;  55-164
Barone, J., 54-176
Barrett, K. W., 57-14
Earth,  D.  S., 54-94
Barth,  E. F., Ill,  55-224
Bartley, R. W., 54-35
Bartolomeo, A. S., 52-156
Baughman, K. J., 52-58
Baughman, W. A., 55-126
Baumwoll, D., 55-22
Baxter, T. A., 54-341
Bayse,  D.  D., 54-253
Beam,  P., 55-84
Beam,  P. M., 57-84; 55-71
Beck, W. W., Jr., 50-135; 52-94; 55-13
Becker, J. C., 55-442
Beckert, W. F., 52^5
Beckett, M. J., 52-431
Beekley, P., 55-97
Beers,  R. H., 57-158
Beilke, P.  J., 52^24
Bell, R. M., 52-183, 448;  54-588
Ben-Hur, D., 54-53
Benson, B. E., 50-91
Benson, J., 55-386
Benson, R. C., 50-59;  57-84; 52-17; 55-71;
  55-112; 55-465
Berger, I.  S., 52-23
Berk, E., 55-386
Berkowitz, J., 55-301
Bernard, H., 50-220; 55-463
Bernert, J. T., 54-253
Berning, W., 55-386
Bhalla, S., 55-189
Bilello, L. J., 55-248
Billets, S., 54-45
Binder, S., 55-409
Bird, K. J., 55-126
Bissex, D. A., 55-208
Bixler,  D.  B., 52-141;  54-493
Blackman, W. C., Jr., 50-91; 54-39; 55-407
Blais, L., 55-441
Blasland, W. V., Jr., 57-215; 55-123
Blayney, E.  K. H., 55-476
Boa, J. A., Jr., 52-220
Bogue, R. W., 50-111
Bonazountas, M., 54-97
Bond,  F. W., 52-118
Bonneau, W. F., 54-509
Boornazian, L. Y., 55-398
Bopp,  F.,  Ill, 54-176
Borsellino, R. J.,  55-299
Bort, R. M., 55-152
Bouck, W. H., 57-215
Boutwell, S. H., 55-135
Bove, L. J., 54-412
Bowders, J. J., 57-165
Boyd,  J., 54-382
Boyd, K. A., 55-61
Bracken, B. D., 52-284
Bradford, M. L.,  52-299
Bradley, C. K., 55-120
Bradshaw, A. D., 52-183
Bramlett, J. A., 55-237
Brandwein, D. I., 50-262; 57-398
Brandwein, S. S., 52-91
Brannaka, L. K.,  57-143
Braun, J. E., 54-449
Brenneman, D., 55-299
Bridges, E. M., 54-553
Brink, J. M., 54-445, 504
Brockbank, B. R., 54-371
Brodd, A. R., 52-268
Brokopp, C., 54-239
Brown,  K. W., 57-223; 54-94, 185; 55-442
Brown,  M. J., 52-363
Brown,  S. M., 57-79; 55-135
Bruck, J., 55^52
Bruck, J. M., 54-72
Bruehl,  D. H., 50-78
Brugger, J. E., 50-119, 208; 57-285; 52-12
Brunner, P. G., 55-43
Brunotts, V. A., 55-209
Brunsing, T. P., 52-249; 54-135
Bryson,  H. C., 50-202
Buechler, T. J., 55-61
Buecker, D. A., 52-299
Buehler, R., 55-208
Buelt, J. L., 54-191
Buhts, R. E., 55-456
Buller, J., 55-395
Bumb, A. C., 54-162
Burgess, A. S., 55-331
Burgher, B. J., 52-357; 54-335
Burns, H., 55-428
Burruel, J. A., 55-318
Burrus,  B. G., 52-274
Burse, V. W., 54-243
Bush, B., 50-212
Butler, H. P., 52-418
Butterfield, W. S., 52-52
Buttich, J. S., 54-200
Byers, W. D., 54-170
Byrd, J. F., 50-1

Cadwallader,  M. W., 55-282
Caldwell, S., 57-14
Campbell, D. L.,  55-36
Campbell, P. L., 54-145
Cane, B. H.,  52-474
Caravanos, J., 54-68
Carnow, B. W., 55-455
Carson, L.  P., 56-445
Carter, J.  L., 55-192
Carter,!. D., 55-63
                                                                                                              AUTHOR INDEX     551

-------
       Casteel, D., 50-275
       Castle, C., 55-452
       Cavalli, N. J.. 54-126
       Celender, J. A., 57-346
       Chaconas, J. T., 57-212
       Chan.R.,  55-98
       Chang, R., 55-97
       Chang, S. S., 5M4
       Chapman, G. H., 5(5-120
       Chappell, R. W., 56-115
       Chamley, G., 56-193
       Chase, D. S., 55-79
       Cheatham, R. A., 56-386
       Chieh, S-H, 84-\
       Childs, K. A., 52^437
       Chisholm, K., 56-420
       Cho, Y.,  83-420
       Christofano, E.  E., 50-107
       Christopher, B.  R., 56-247
       Christopher, M. T.,  50-233
       Chrostowski, P. C.,  56-242
       Chung, N. K., 50-78
       Cibulas, W., Jr., 56-467
       Cibulskis, R. W..  52-36
       Cichowicz, N, L., 50-239
       Clark, R., 54-486
       Clarke, J. H., 55-296
       Clay, P. F.. 57-45, 52-40; 55-100; 56-120
       Clemens,  B., 54-49, 335; 55-419; 56-445
       Clemens,  R., 54-213
       demons,  G. P., 54-404
       Cline, P.  V., 54-217
       Clinton, R. J., 56-4
       Coatcs, A. L., 56-365
       Cochran,  S. R.,  52-131; 54-498
       Cochran,  S. R.,  Jr.,  50-233; 55-275
       Cogliano. V. J., 56-182
       Cohen, S. A., 5/-405
       Coia, M.  F.. 56-322
       Coldeway, W. G., 54-584
       Cole, C. R., 5/-306;  52-118
       Collins, J. P., 5/-2; 55-326
       Collins, L. 0., 55-398
       Colonna.  R., 50-30
       Conibear, S. A., 56-455
       Cook, D.  K., 5/-63
       Cook. L.  R., 55-280
       Cooper, C., 5/-185
       Cooper, D., 55-419;  56^457
       Cooper, E. W., 55-338
       Cooper, J. W.. 52-244
       Cooper, L. M., 56-415
       Copeland, L. G., 56-287
       Corbett, C.  R., 50-6; 5/-5
       Corbin, M.  H., 56-322
       Corbo,  P., 52-203
       Corn, M.  R., 5/-70
       Cornaby,  B. W., 52-380
       Cothron, T. K...  54-452
       Coutre, P. E., 54-511
       Cox, G. V., 5/-1
       Cox, R. D., 52-58, 334
       Crawford, R. B., 56-272
       Crawley, W. W., 54-131. 185; 55-80
       Crosby, T. W., 56-258
       Cudahy, J. J., 55-460
       Cullinane, M. J., Jr., 54-465
       Curry, J., Jr., 54-103
       Curry, M. F. R., 56-297
       Czapor, J. V., 54-457

       Dahl, T. 0., 5/-329
       Daigler, J., 55-296
       Daily,  P.  L., 55-383
       Dalton, D. S., 55-21
Dalton, T. F., 5/-371
Davey. J. R., 50-257
Davis, A. 0.. 56-115
Davis, B. D., 54-213
Davis, L. R., 56-303
Davis, S. L., 54-449
Dawson, G. W., 5/-79; 52-386;
  55-453; 56-173
Day,  A.  R., 55-140
Day,  S. R., 56-264
DeCarlo, V.  J., 55-29
Deck, N., 56-38
DeGrood., T. J.. 55-231
Dehn, W. T., 55-313
Deigan, G. J.,56-287
Del Re, S., 56-110
DeLuca,  R. J..  56-148
Demarest, H. I-.., 56-143
Dcmcny, D. D., 56-247
Dcmmy.  R. H., 5/-42
Dempsey, J.  G., 55-26
Denbo, R. T., 56-56
DeRosa,  C.,  55-412
Derrington, D., 54-382
Desmarais, A. M. C., 54-226
Devary, J. L., 55-117
Dhamotharan, D. S., 56-56
Dickinson, R. F., 54-306
Dickinson, R. W.. 56-258
Dickinson, W.,  56-258
DiDomenico, D., 52-295
Diecidue, A.  M., 52-354; 55-386; 56-22
Dienemann, E.  A., 54-393
Dies), W. F.. 50-78
Dikinis. J. A.. 54-170
Dime, R. A., 55-301
DiNapoli. J.  J., 52-150
DiNitto,  R. G., 52-111; 55-130
DiPuccio, A., 52-311
Dirgo, J. A., 56-213
Diugosz,  E. S.,  55-429
Dodge, L., 55-255
Dorrler,  R. C.,  54-107
Dowiak,  M. J., 50-131; 52-187; 54-356
Doyle, D. F., 55-281
Doyle, R. C., 52-209
Doyle, T. J., 50-152
Dragun,  J., 56-453
Drake, B., 52-350
Drever, J. I., 54-162
Dnscoll,  K.  H., 5/-103
Duff. B.  M., 52-31
Duffala.  D. S.,  52-289
Duffee, R. A., 52-326
Duke, K.  M., 52-380
Duncan,  D., 5/-21
Dunckel, J. R., 55-468; 56-361
Du  Pont, A., 56-306
Dursi. C. M., 55-234
Duvel, W. A., 52-86
Dybevick. M. H., 55-248

Earp, R.  F.. 52-58
Eastman, K.  W., 55-291
Eastwood, D., 56-370
Eberhardt. L. L., 54-85
Eckel, W. P., 54-49; 55-130
Edwards.  J. S., 55-393
Ehrlich, A. M.. 56-167
Ehrman,  J., 54-374
Eicher, A. R., 55-460
Eimutis,  E. C.,  5/-123
Eissler, A. W.,  54-81
Eklund, B. M.,  54-77
Eley,  W.  D., 54-341
Elkus, B.. 52-366
Ellis, H. V., Ill, 56-213
Ellis, R. A., 52-340
Eltgroth, M. W., 55-293
Emerson. L. R., 55-209
Emig, D. K., 52-128
Emmett, C. H., 56-467
Emrich. G.  H.,  50-135; 56-412
Eng, J., 54-457
Engels, J. L., 54-45
Englcr. D. R., 55-378
English, C.  J., 55-453; 54-283; 56-173
Englund. E.  J..  56-217
Erbaugh, M., 55-452
Erdogan, H, 55-189
Esposito, M. P., 54-486; 55-387
Ess, T., 52-390,408
Ess. T. H.,  5/-230
Evans,  J. C., 52-175;  55-249,357,369
Evans.  M.  L., 54-407
I:vans,  R. B., 52-17; 55-28
Evans,  R. G., 56-78
Evans,  T. T, 54-213
Everett, L. G., 52-100
Exner, P. J.. 54-226

Fagliano, J. A.,  54-213
Falcone, J. C., Jr., 52-237
Falk, C. D., 56-303
Fang. H-Y,  52-175; 55-369
Farrell, R. S.. 55-140
Farro, A., 55-413
Fast, D. M.. 54-243
Faulds, C. R., 54-544
Fold. R. H., 55-68
Fell. G. M., 55-383
FeUows, C.  R., 55-37
Fenn, A. H., 55-476
Ferenbaugh,  R.  W.. 56-1
Ferguson, J., 54-248
Ferguson, T., 50-255
Fields, S . 54-404
Figueroa, E.  A.. 5/-313
Fine, R. J.,  54-277
Finkbeiner,  M. A., 55-116
Finkel,  A. M., 5/-341
Fischer, K. E., 50-91
Fisher.  W  R.. 56-124
Fisk. J. F., 55-130
Fitzpatrick,  V. F..  54-191; 56-325
Flatman, G. T.,  55-442; 56-132,  217
Ford. K. L., 54-210. 230
Forney, D.,  55-409
Forrester, R., 57-326
Fortin.  R. L.. 52-280
Foth, D. J., 56-176
Francingues, N.  R., 52-220
Franconeri,  P., 5/-S9
Frank, J.. 54-532
Frank, U., 50-165; 5/-96,  110
Fredericks, S., 56-36,  120
Freed, J. R., 50-233
Freestone, F. J., 50-160, 208; 5/-28S
Freudenthal, H.  G.. 52-346
Friedman, P. H.. 54-29. 49
Friedrich, \V., 55-169
Fuller, P. R., 56-313
Funderburk.  R., 54-195
Furman, C., 52-131
Furst, G. A., 55-93

Galbraith, R. M., 56-339
Gallagher, G. A., 50-85
Galuzzi, P.. 52-81
Garczynski,  L., 54-521; 56-40
552     AUTHOR INDEX

-------
Garlauskas, A. B., 83-63
Garnas, R. L., 84-39
Garrahan, K. G.(  54-478; 86-167
Gay, F. T., Ill, 52-414
Geil, M., 55-345
Geiselman, J. N.,  55-266
Gemmill, D., 55-386; 54-371
Gensheimer,  G. J., 54-306
George, J. A., 5(5-186
Geraghty, J.  J., 50-49
Geuder, D., 54-29
Ghassemi, M., 50-160
Ghuman, 0. S., 54-90
Gianti, S. J.,  54-200
Gibbs, L. M., 55-392
Gibson, S. C., 57-269
Gigliello,  K., 54-457
Gilbert, J. M., 52-274
Gilbertson, M. A., 52-228
Gill, A., 54-131
Gillen, B. D., 52-27; 55-237
Gillespie,  D.  P., 50-125; 57-248
Gilrein, S. A., 55-158
Gish, B. D.,  54-122
Givens, R. C., 55-31
Glaccum, R.  A., 50-59; 57-84
Glynn, W. K., 56-345
Goggin, B., 57-411
Gold, J.,  54-416
Gold, M. E.,  57-387
Goldman, L. M., 54-277
Goldman, R. K., 57-215
Goldstein, P., 55-313
Golian, S. C., 56-8
Goliber,  P., 50-71
Golob, R. S., 57-341
Golojuch, S.  T., 55-423
Goltz, R. D., 52-262; 55-202;
  54-489; 55-299
Goode, D. J., 55-161
Goodman, J., 55-419
Goodwin, B.  E., 55-7
Gorton, J. C., Jr., 57-10; 54-435
Goss, L. B.,  52-380
Gray, E. K.,  55-406
Graybill, L.,  55-275
Greber, J. S. 84-486; 55-387
Green, J., 57-223
Greiling, R. W., 54-535
Griffen, C. N., 55-53
Grube, W. E., Jr., 52-191, 249
Gruenfeld, M., 50-165; 57-96;  52-36
Guerrero, P., 55-453
Gurba, P., 54-210, 230
Gurka, D. F., 52-45
Gushue, J. J., 57-359; 55-261
Gustafson, M. E.,  5(5-448
Guthrie,  J., 56-386

Hadzi-Antich, T.,  56-18
Haeberer, A. F., 52-45
Hagel, W. A., 56-186
Hager, D. G., 52-259
Hagger, C., 57-45; 54-321; 55-7
Hahn, S.  J.,  56-448
Haji-Djafari, S., 55-231
Hale, F.  D.,  55-195
Halepaska, J. C., 54-162
Hall, J. C., 54-313; 56-27
Hallahan, F.  M., S5-14
Haller, P. H., 56-469
Hammond, J. W., 50-250; 57-294
Hanley, M. M., 52-111
Hansel, M. J., S5-253
Hanson,  B., 52-141;  55-4; 56-224, 462
Hanson, C. R., 54-189; 55-349
Hanson, J. B., 57-198; 54-493
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; 54-363
Hatch, N. N., Jr., 55-285
Hatheway, A.  W., 55-331
Hawkins, C., 55-395
Hawley, K. A., 55-432
Hayes, E.,  55-285
Hazaga, D., 54-404
Head, H. N.,  56-258
Heare, S., 55-395
Hediger, E. M., 56-164
Heeb, M., 57-7
Heffernan, A. Z.,  56-8
Hemsley, W. T., 50-141
Henderson, D. R., 56-380
Henderson, R. B., 54-135
Hendry, C. D., 55-314
Hennington, J. C., 55-21; 55-374
Hess, J. W., 55-108
Heyse, E.,  55-234
Hijazi, N., 55-98
Hilker,  D., 50-212
Hill, J.  A., 56-292
Hill, R., 52-233
Hill, R. D., 50-173; 56-356
Hillenbrand, E.,  52-357, 461
Hina, C. E., 55-63
Hines, J. M.,  S7-70;S5-349
Hinrichs, R., 50-71
Hinzel,  E. J.,  56-313
Hirschhorn, J. S.,  55-311
Hitchcock, S., 52-97; 56-318
Hjersted, N. B., 50-255
Hoag, R. B., Jr., 55-202
Hodge,  V., 54-62,  498
Hoffman, R. E., 56-78
Holberger,  R.  L., 52-451
Holmes, R. F., 54-592
Holstein, E. C., 54-251
Homer,-D. H., 56-213
Hoogendoorn, D., 54-569
Hooper, M. W., 55-266
Hopkins, F., 50-255 ,
Home,  A., 57-393
Horton, K. A., 57-158
Hosfeld, R. K., 56-415
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
Hubbard, A. E., 56-186
Hubbard, R. J., 56-186
Huffman, G. L., 54-207
Hughey, R. E., 55-58
Huizenga, H.,  55-412
Hullinger, J. P.,  55-136; 55-158
Hunt, G. E., 50-202
Hunter,  J. H., 55-326
Hupp, W. H., 57-30
Hutson, K. A., 56-8
Hwang,  J. C., 57-317;  54-1
Hwang,  S.  T., 54-346

laccarino, T.,  54-66
Ing,  R., 54-239
Ingersoll, T. G., 57-405
 Ingham, A. T., 55-429
 Isaacson, L., 57-158
 Isaacson, P. J., 55-130
 Isbister, J. D., 52-209
 Iskandar, I. K., 54-386

 Jacobs, J. H.,  52-165
 Jacobson, P. R., 56-233
 Jacot, B. J., 55-76
 James,  S. C., 50-184; 57-171, 288; 52-70
   131; 54-265;  55-234
 Janis, J. R., 57-405; 52-354
 Janisz,  A. J., 52-52
 Jankauskas, J.  A., 55-209
 Jarvis, C. E., 54-469
 Jerrick, N. J.,  55-389; 54-368
 Jessberger, H.  L., 55-345
 Jhaveri, V., 55-242; 55-239
 Johnson, D., 54-544
 Johnson, D. W., 56-227
 Johnson, G. M., 56-93,  105
 Johnson, M. F., 56-52
 Johnson, M. G., 57-154
 Johnson, W. J., 56-227
 Johnson-Ballard, J.,  57-30
 Johnston, R. H., 55-145
 Jones, A. K., 52-183, 448
 Jones, B., 54-300; 55-412, 419
 Jones, K. H., 52-63
 Jones, R. D., 55-123, 346
 Jones, S. G., 55-154
 Jordan, B. H.,  52-354
 Jowett,  J. R., 54-339; 56-40
 Jurbach, R., 54-66

 Kaczmar, S. W., 54-221
 Radish, J., 52-458
 Kaelin,  J. J., 55-362
 Kaltreider, R., 56-14, 398
 Kaplan, M., 52-131
 Karably, L. S.,  56-436
 Karon, J. M., 54-243
 Kaschak, W. M., 52-124; 54-440;
  55-281; 56-393
 Kastury, S., 55-189
 Katz, S., 55-419
 Kay, R. L., Jr., 54-232
Kay, W.,  55-409
Keffer, W., 54-273
Keirn, M. A., 55-314
Keitz, E. L., 52-214
Kemerer, J. A., 54-427
Kennedy,  S. M., 57-248
Kerfoot, H. B., 54-45
Kerfoot, W. B., 57-351
Khan, A.  Q., 50-226
Khara, B. H., 56-220
Kilpatrick, M. A., 50-30; 54-478
Kim,  C. S., 50-212
Kimball, C. S.,  55-68
King, J., 54-273; 55-243
Klinger, G. S.,  55-128
Knowles, G. D., 55-346
Knox, J. N., 56-233
Knox, R. C., 55-179
Koerner, R. M., 50-119; 57-165, 317;
  52-12; 55-175; 54-158; 56-272
Koesters, E. W., 54-72
Kolsky,  K., 54-300
Kopsick, D. A., 52-7
Kosin, Z., 55-221
Kosson, D. S.,  55-217; 54-393
Koster, W. C.,  50-141
Koutsandreas, J. D.,  55-449
Kovell, S. P., 56-46
                                                                                                           AUTHOR INDEX     553

-------
      Kraus,  D. L., 55-314
      Krauss, E. V., 86-138
      Kufs, C.  T., 80-30; 82-146; 86-110
      Kugelman, I. J., 55-369
      Kunce, E. P., 86-345
      Kuykendall,  R.  G., 83-459

      LaBar, D., 85-449
      LaBrecque, D., 83-26
      Lacy, W. J., 84-592
      LaFornara, J. P., 87-110, 294; 85-128
      LaGrega, M. D., 8/-42
      LaMarre, B. L., 82-291
      Lambert, W. P.. 8*412
      Lamont,  A., 84-16
      Lampkins, M. J., 85-318
      Langner, G., 82-141
      Lappala, E.  G., 84-20
      Larson, R. J., 80-180
      Laswell, B. H., 85-136
      Lataille, M., 82-57
      Lavinder, S. R., 85-291
      Lawrence, L. T., 84-481
      Lawson,  J. T., 82-474
      LeClare,  P.  C., 85-398
      Lederman.*  P.  B., 80-250; 87-294
      Lee, C. C., 82-214; 84-207
      Lee, G. W., Jr., 85-123, 346
      Lee, R. D., 85-157
      Leighty, D. A., 85-79
      Ltis, W.  M.. 80-116
      Lennon, G.  P., 85-357
      Leo, J., 82-268
      Leu, D. J., 86-303
      Lewis,  D. S., 84-382
      Lewis,  W. E., 84-427
      Lidberg,  R., 86-370
      Liddle, J. A., 84-243
      Lincoln, D. R., 85-449
      Linkenheil, R., 85-323
      Lippe,  J. C., 85-423
      Lippitt, J. M., 82-311; 85-376
      Lipsky, D., 82-81
      Livolski,  J. A., Jr.,  84-213
      Lo, T.  Y. R., 85-135
      Lombard, R. A., 85-50
      Longstreth, J., 85-412
      Lord, A. E., Jr., 80-119; 87-165; 82-12;
        85-175; 84-158; 86-272
      Losche, R., 87-96
      Lough, C. 3., 82-228
      Lounsbury, J., 84-498; 86-457
      Loven, C. G., 82-259
      Lowe, G. W., 84-560
      Lowrance, S. K., 85-1
      Lucas,  R. A., 82-187
      Lueckel, E. B.,  85-326
      Lundy, D. A., 82-136
      Lunney, P.,  82-70
      Lybarger, J.  A., 86-467
      Lynch,  D. R., 84-386
      Lynch,  E. R., 87-215
      Lynch,  J. W., 80-42; 85-323
      Lysyj, I., 87-114; 85^46

      Mack, J., 84-107
      MacRoberts, P. B., 82-289
      Magee, A. D., 85-209
      Mahan, J. S.,ฎ  82-136
      Makris, J., 86-11
      Malone, P. G.,  80-180; 82-220
      Maloney, S.  W., 85-456
      Mandel, R. M., 80-21
      Manko, J. M., 87-387
      Mann,  M. J., 85-374
Manuel, E. M., 85-249
Margolis,  S., 85-403
Marshall,  T. C., 84-261
Martin. W. ?., 85-322; 84-248
Martin. W. J., 82-198; 86-277
Martz, M. K., 86-1
Maser. K. R., 85-362
Mashni, C. 1., 86-237
Maslansky, S. P., 82-319
Maslia, M. L.. 85-145
Mason, B. J., 84-94
Mason, R., 86-52
Mason, R. J.. 84-339
Massey. T. I.. 80-250
Mateo. J., 86-14
Mateo, M.. 85-413
Mathamel, M. S., 8/-280; 86-472
Matthews, R. T., 85-362
Maughan, A. D.. 84-239
Mavraganis, P. J., 85-449
Mazzacca, A. J., 85-242; 85-239
McAneny. C. C., 85-331
McArdle,  J., 84-486
McCloskey,  M. H., 82-372
McClure,  A. F., 84-452
McCord, A. T., 87-129
McCracken, W. E., 86-380
McEnery.  C. L., 82-306
McCarry,  F. J.. 82-291
McGinnu, J. T., 82-380
McGlew. P.  J., 84-150; 85-142; 86-403
McGovern, D., 84-469
McKee. C. R., 84-162
McKown,  G. L., 87-300, 306; 84-283
McLaughlin, D. B.. 80-66
McLeod, D. S., 84-350
McLeod, R.  S., 84-114
McMillan, K. S., 85-269
McMillion, L. G., 82-100
McNeill, J. D., 82-1
Meade, J. P.,  84-407
Mehran, M., 85-94
Meier, E.  P., 82-45
Melvold, R.  W., 87-269
Menke, J. L.,  80-147
Mercer, J. W., 82-159
Mernitt, S.,  85-107
Messick, J. V., 87-263
Messinger, D.  J., 86-110
Meyer, J., 80-275
Meyers, T. E., 80-180
Michelsen, D.  L., 84-398; 85-291
Milbrath,  L. W., 87-415
Militana. L.  M., 86-152
Miller, D. G.,  Jr., 82-107; 85-221
Miller, K.  R., 85-136;  86-158
Mindock,  R. A.. 86-105
Mitchell, F. L., 84-259; 85-406
Mittleman, A. L., 84-213
M onsets, M., 85-88
Monserrate,  M., 86-14
Montgomery. R. J., 86-292
Montgomery, V. J., 85-8
Mooney, G.  A., 84-35
Moore, S. F.. 80-66
Morahan.  T. J.. 85-310
Moran. B. V., 85-17
Morey, R. M., 87-158
Morgan, C. H., 80-202
Morgan, R. C., 82-366; 84-213; 85-396
Morin. J.  0., 85-97
Morson, B. J., 84-535
Mortensen, B. K., 86-74
Morton, E. S., 86-213
Moscati, A.  F., Jr., 86-164, 420
Moslehi, J., 85-326
Mott, R. M., 80-269; 85-433
Motwani, J. N., 86-105
Mousa, J. J., 85-86
Moyer,  E. E., 85-209
Moylan. C.  A., 85-71
Muller,  B. W., 82-268
Muller-Kirchenbauer, H., 85-169
Mullins, J. W.. 85-442
Munoz, H., 84-416
Murphy, B.  L., 82-331, 3%; 85-13
Murphy, C. B., Jr.,  85-195; 84-221
Murphy, J.  R., 84-213
Murray, J. G., 85^*64
Musser, D. T.. 85-231
Mulch,  R. D., Jr., 85-296
Myers. V. B.,  82-295; 85-354
Myrick, J., 84-253

Nadeau, P. F.. 82-124; 85-313
Nadeau. R. J., 85-128
Nagle, E., 85-370
Narang, R., 80-212
Naugle, D. F., 85-26
Nazar, A.. 82-187; 84-356
Needham, L. L., 84-253; 86-78
Neely, N S., 80-125
Nelson.  A.  B.. 87-52
Nelson,  D. D., 85-32
Neumann. C.,  82-350
Newman, J. R., 84-350
Newton. C. E.. 86-420
Nichols, F. D.. 84-504
Nickens, D.. 84-416
Nielsen, D. M . 86-460
Nielsen, M . 87-374
Niemele, V.  E., 82-437
Nimmons, M  J.,  85-94
Nisbet, I. C. T.. 82-406
Noel, M. R.. 85-71
Noel, T. E.. 85-266
Noland. J. \V., 84-P6. 203
Norman, M  , 86-318
Norman, \V  R., 82-111; 85-261
North, B. E.. 87-103
Nyberg, P. C., 84-504
Nygaard, D. D., 85-79

Obaseki. S., 84-598
O'Dea, D., 85-331
O'Flaheny. P. M  , 84-535
Ogg, R. N .  85-202, 358; 86-356
O'Hara, P. F., 86-126
Ohonba, E., 84-598
Oi. A. W., 87-122
O'Keefe, P., 80-212
Okeke, A. C..  85-182
Olsen, R. L., 85-107; 86-115,  31.1, 386
Opitz, B. E., 86-277
Oma, K. H., 84-191
O'Malley, R., 85-58
Openshaw, L-A, 85-326
Opitz, B. E., 82-198
Orr, J. R., 85-349
Ortiz, M., 86-84
Osborn. J., 85-43
Osheka, J. W., 80-184
Oster, J. G., 86-138
O'Toole, M. M., 85-116
Ounanian. D. W.,  85-270
Owens,  D. W., 80-212

Page, R. A., 84-594
Paige, S. F., 80-30, 202
Palombo, D. A., 82-165
554     AUTHOR INDEX

-------
Pajak, A. P., 50-184; 87-288
Paquette, J. S., 56-208, 393
Parks, G. A., 83-280
Parker, F. L., 87-313
Parker, J.  C., 54-213
Parker, W. R., 54-72
Parratt, R. S.,  55-195
Parrish, C. S.,  55-1
Parry, G. D. R., 82^48; 54-588
Partridge, L. J., 54-290; 55-319; 56-65
Partymiller, K., 54-213
Paschal, D., 55-409
Paschke, R. A., S5-147
Patnode, T.,  55-323
Patterson, D. G., Jr., 56-78
Pearce, R. B., 57-255; 55-320
Pearsall, L. J., 56-242
Pease, R. W., Jr., 50-147; 57-171, 198
Pedersen, T. A., 56-398
Pennington, D., 55-253
Peters, J. A., 57-123
Peters, N., II, 56-365
Peters, W. R., 52-31
Peterson, J. M., 55-199
Phillips,  J. W., 57-206
Pickett, J.  S., 56-424
Pierson, T., 54-176
Pintenich, J.  L., 57-70
Plourd, K. P., 55-396
Plumb, R. H.,  54-45
Ponder, T. C., 55-387
Popp, S. A., 56-105
Possidento, M., 50-25
Possin, B.  N., 53-114
Powell, D. H., 53-86
PredpaU, D. F., 54-16
Preston, J. E.,  54-39
Preuss, P.  W.,  56-167
Price, D. E., 54-478
Price, D. R., 52-94
Priznar, F. J., 55-1,  74; 56-84
Proko, K., 55-11
Prothero, T. G., 54-248
Prybyla, D. A., 55-468
Pyles, D. G., 56-350

Quan, W., 57-380
Quimby, J. M., 52-36
Quinlivan, S., 80-160
Quinn, R. D., 86-393
Quinn, K. J., 84-170; 55-157
55-157
Quintrell, W. N., 55-36

Rademacher, J. M.,  54-189; 85-349
Rams, J. M., 87-21
Ramsey, W. L., 80-259; 57-212
Ransom, M., 50-275
Rappaport, A., 57-411
Rea, K. H., 86-1
Rebis, E. N., 53-209
Reifsnyder, R. H., 52-237
Reiter, G. A., 50-21
Remeta, D. P., 80-165;  87-96
Repa, E., 52-146; 85-164
Repa, E. W., 86-237
Reverand, J.  M., 54-162
Rice,  E. D., 55-84
Rice,  J. M., 85-182
Rice,  R.  G., 84-600
Richards, A., 80-212
Richardson, S., 54-1
Rick,  J., 54-469
Ridosh, M. H., 84-427; 85-243
Rikleen,  L. S.,  82-470; 55-275
Ritthaler, W. E., 52-254
Riner,  S. D., 52-228
Rishel, H. L., 57-248
Rizzo,  J.,  82-17
Rizzo,  W. J., Jr., 85-209
Robbins, J.  C., 83-431
Roberts, B.  R., 83-135
Roberts, D.  W., 86-78
Rockas, E.,  85-11
Rodricks, J. V., 53-401
Rogers, W., 54-16
Rogoshewski, P. J.,  50-202; 82-131,
  146;  54-62
Romanow, S., 85-255
Ronk,  R. M., 86-471
Roos, K. S., 53-285
Rosasco, P.  V., 84-103
Rosbury, K. D., 84-265
Rosebrook,  D.  D., 84-326
Rosenkranz, W., 87-7
Ross, D., 54-239
Rothman,  D. W., 54-435
Rothman,  T., 52-363
Roy, A. J.,  83-209
Royer,  M. D., 87-269
Rubenstein,  P.  L., 86-143
Ruda,  F. D., 84-393
Rulkens, W. H., 52-442;  54-576
Rupp,  M.  J., 86-164
Ryan, C. R., 56-264
Ryan, F. B., 87-10
Ryan, M.  J., 85-29
Ryan, R. M., 55-125
Ryckman, M. D., 54-420

Sadat,  M. M., 53-301, 413
Salvesen, R. H., 54-11
Sanders, D.  E., 52-461
Sandness,  G. A., 57-300; 53-68
Sandza, W.  F., 55-255
Sanford, J. A., 54-435
Sanning, D.  E., 57-201; 82-118, 386
Sappington,  D., 85-452
Sarno,  D.  J., 85-234
Schafer, P. E., 85-192
Schalla, R.,  83-117; 54-283
Schaper, L.  T., 56-47
Schapker,  D. R., 56-47
Schauf, F. J.,  50-125
Scheppers, D.  L., 84-544
Schilling, R., 84-239
Schlossnagle, G. W., 53-5,  304
Schmidt, C. E., 52-334; 53-293
Schnabel, G. A., 50-107
Schneider, P., 50-282
Schneider, R., 50-71
Schnobrich,  D. M., 55-147
Schoenberger,  R. J., 52-156
Schofield, W. R., 54-382
Scholze, R.  J.,  Jr.,  55-456
Schomaker,  N. B., 50-173;  52-233
Schuller, R.  M., 52-94
Schultz, D. W., 82-244
Schweitzer, G. E., 87-238; 52-399
Schweizer, J. W., 56-339
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
Sebastian, C.,  86-14
Sebba, F., 84-398
Segal, H. L., 85-50
Selig, E. I.,  82-458; 53-437
Sepesi, J. A., 55-423, 438
Sevee, J. E., 52-280
Sewell, G.  H., 82-76
Seymour, R. A., 82-107
Shafer, R.  A., 84-465
Shapot, R. M., 86-93
Sharkey, M.  E., 54-525
Sharma, G. K.,  87-185
Shaw, E. A., 86-224
Shaw, L. G., 87-415
Sheedy, K. A., 80-116
Shen, T. T.,  52-70, 76; 54-68
Sheridan, D. B.,  54-374
Sherman, J.  S.,  52-372
Sherwood,  D. R., 82-198; 86-277
Shih, C. S., 57-230; 52-390, 408; 53-405
Shimmin, K. G., 56-143, 463
Shiver, R. L., 55-80
Shoor, S. K., 56-4
Shroads, A. L.,  53-86
Shuckrow,  A. J., 50-184; 57-288
Shugart, S. L., 56-436
Shultz, D. W., 82-31
Sibold, L. P., 55-74
Siebenberg, S., 54-546
Silbermann, P. T.,  50-192
Silcox, M. F., 83-8
Silka, L. R.,  80-45; 82-159
Sills, M. A.,  80-192
Simanonok, S. H.,  86-97
Simcoe, B., 57-21
Simmons, M. A., 54-85
Sims, R. C.,  53-226
Singer, G. L., 54-378
Singh, J., 84-81
Singh, R., 53-147
Sirota, E. B., 53-94
Siscanaw, R., 52-57
Sisk, W. E.,  54-203, 412
Skalski, J. R., 54-85
Skoglund, T. W., 55-147
Slack,  J., 50-212
Slater, C. S., 52-203
Sloan, A., HI, 55^38
Slocumb, R.  C., 56-247
Smart, R. F., 54-509
Smiley, D., 54-66
Smith, C., 54-546
Smith, E. T., 80-8
Smith, J. S.,  84-53
Smith, L. A., 85-396
Smith, M. A., 82-431; 84-549
Smith, M. 0., 86-430
Smith, P., 56-313
Smith, R., 50-212
Smith, R. L., 85-231
Smith, S., 86-462
Smith, W., 86-333
Snow, M., 85-67
Snyder, A.  J., 87-359
Snyder, M., 50-255
Sokal, D., 54-239
Solyom, P., 83-342
Sosebee, J. B., 84-350
Sovinee, B., 55-58
Spatarella, J. J., 54-440
Spear, R., 57-89
Spencer, R. W.,  52-237
Spittler, T. M., 57-122; 52-40, 57;
  53-100, 105; 85-93
Spooner, P. A.,  80-30, 202; 52-191;
  85-214, 234
Springer, C., 52-70
Springer, S. D.,  56-350
Srivastava, V. K., 53-231
Staible, T., 85-107
                                                                                                            AUTHOR INDEX     555

-------
   Stammler,  M., 83-68
   Stanfill, D. F., Ill, 85-269
   Stanford, R. L.. 57-198; 84-498; 85-275
   Stankunas, A.  R., 52-326
   Stanley, E. G., 85-1
   Starr, R. C., 80-53
   St. Clair, A. E., 82-372
   Steele, J. R., 84-269
   Steelman, B. L., 85-432
   Stehr, P.  S., 84-287
   Stehr-Green, P. A.. 86-78
   Steimle,  R. R., 87-212
   Stein, G. F., 84-287
   Steinberg, K.  K., 84-253
   Stephens, R. D.,  80-15; 82-428; 85-102
   Stief. K., 82-434; 84-565
   Stokely, P. M., 84-6
   StoUer, P. J.. 80-239; 87-198
   Stone, T., 85-128
   Stone, W. L.,  87-188
   Stoner, R., 84-66
   Strandbergh, D., 84-81
   Strattan,  L. W.. 87-103
   Strauss, J. B., 87-136
   Slrenge, D. L., 85-432
   Strickfaden, M. E., 85-7
   Strong, T. M., 85-473
   Stroud, F. B.. 82-274
   Stnizziery, J. J., 80-192
   Sullivan.  D. A., 87-136
   Sullivan, J. H., 83-31
   Sullivan, J. M., Jr., 84-386
   Sutton, P. M.. 86-253
   Swaroop, A., 84-90
   Swatek, M. A., 85-255
   Swenson, G. A., Ill, 85-123
   Swibas, C. M., 84-39
   Sydow, W. L., 86-393,  398

   Tackett. K. M, 87-123
   Tafuri, A. N., 87-188; 82-169; 84-407
   Tanzer, M. S., 87-10
   Tapscott, G., 82-420
   Tarlton, S. F., 8*445
   Tate, C. L.. Jr.,  84-232
   Taylor, B., 85-304
   Taylor, M. D., 86-88
   Teets, R. W., 85-310
   Teller. J., 84-517
   Tewhey, J. D., 82-280; 84-152
   Thiesen. H. M., 82-285
   Thibodeaux, L. J., 82-70
   Thomas, A., 84-176
   Thomas, C. M., 85-112
   Thomas,  G. A., 80-226
   Thomas, J. E., Jr., 84-150; 85-142
   Thomas,  J. M., 84-85
   Thomas,  S. R., 85-476
   Thompson, G. M., 84-20
   Thompson, S. N.( 85-331
   Thompson, W. E.. 84-469; 85-387
   Thomsen, K.  0., 86-138. 220
   Thorsen. J. W., 87-42,  259; 82-156
   Thorslund, T. W., 86-193
    ThrelfaU, D., 80-131; 82-187
    Tifft, E.  C., Jr., 84-221
Tillinghast, V., 85-93
Timmerman, C. L., 84-191
Tinto. T., 85-243
Titus, S. E., 87-177
Townsend, R. W., 82-67
Trees, D. P.. 8449
Tremblay, J. W., 85-423
Trezek,  G. J.. 86-303
Triegel.  E. K.,  85-270
Troxler,  W. L., 85-460
Tniett. J. B., 82-451
Tucker, W.  A., 84-306
Tuor, N. R., 85-389; 84-368
Turner, J. R., 85-17
Turnham, B., 85-423
Turoff.  B., 80-282
Turpin,  R. D., 87-110, 277; 85-82;
  84-81. 273
Tusa, W. K.. 87-2; 82-27
Twedell, A. M., 80-233
Twedell, D. B., 80-30, 202
Tyagi, S., 82-12
Tyburski, T. E , 85-3%

Unites,  D. F., 80-25; 87-398; 85-13
Unterberg, W., 87-188
Urban.  M. J., 84-53
Urban,  N. W., 82414; 85-5, 304

Vais, C., 84-427
Vanderlaan, G. A., 87-348. 82-321; 85-366;
86-407
VanderVoort, J. D., 86-269
Vandervort, R., 87-263
Van Ee, J. J.,  85-28
van  Epp, T. D., 86-361
Van Gemen, W. J. Th.. 82-442
Van Slyke, D.. 85-442
Velez. V. G.. 86-93
Vias, C., 84-273
Viste, D. R., 84-217
Vocke,  R. W., 86-1
Vogel, G. A., 82-214
von  Braun,  M. C.. 86-200
von  Lindern, I.. 86-31,  200
Voorhees. M. L.. 85-182
Vora, K. H.. 84-81
Vrable, D. L.,  85-378

Wagner, J., 84-97
Wagner, K., 82-169; 85-226; 84-62; 85-221
Walker, K.  D.. 84-321
Wallace, L. P., 85-322
Wallace, J. R., 85-358
Waller, M.  J., 85-147
Wallis,  D. A.,  84-398; 85-291
Walsh,  J., 82-311
Walsh,  J. F., 82-63
Walsh,  J. J., 80-125; 87-248; 85-376
Walther, E. G., 85-28
Wardell, J.. 87-374
Warner, R. C., 86-365
Wasser, M. B., 85-307
Watson. K. S., 85-307
Way, S. C., 84-162
Weaver, R. E. C., 85464
Webb, K. B., 84-287; 86-78
Weber, D. D., 85-28; 86-132, 217
Wciner, P. H., 87-37
Weiss, C., 84-546
Weist, F. C., 85-175
Welks. K. E.,  80-147
Werner, J. D., 85-370; 86-69
Wet/el,  R. S., 80-30, 202; 82-169,  191;
85-234
Wheatcraft,  S.  W., 85-108
Whelan, G., 85-432
White. D. C .  86-356, 361
White, L. A.,  85-281
White, M., 80-275
White, R. M., 82-91
Whitlock, S. A., 85-86
Whitney. H. T., 86436
Whiltaker, K.  F., 82-262
Wiggans, K. E., 85-314
Wilder,  I., 80-173; 82-233
Wiley. J. B., 85-58
Wilkinson, R.  R.,  80-255
Williams,  R. C.. 86-467
Williamson. S. J.. 84-77
Willis. N. M., 86-35
Wilson. D. C., 80-8
Wilson, L. G., 82-100
Wine, J.. 85-428
Winklehauj. C., 85-423
Wirth, P. K.,  84-141
Wise, K. T.. 84-330
Withero*, W. E., 84-122
Witmer, K. A.. 85-357
Wutmann. S  G., 85-157
Woelfel, G.  C.."  85-19:
Wolbach, CD.. 85-54
Wolf. F., 85-43
Wolfe, S. P..  85-88
Wong, J.. 87-374
Woodhouse. D., 85-374
Woodson, L., 86-208
Worden. M  H., 84-273
Worst. N  R., 84-374
\Votherspoon. J.. 86-303
Wright, A. P., 50-42
Wu, B.  C.. 86-350
Wuslich, M  G., a:-224
Wyelh,  R. K , 87-107
Wyman. J.. 85-395

Yaffe, H. J., 80-239
Vang. E. J.. 87-393; 85-370; 84-335; 86-52
Vaniga, P. M., 86-333
Yaohua, Z., 84-604
Yezri, J. J.. Jr., 87-285
Young, L., 80-275
Young, R. A., 87-52
Youzhi, G.. 84-604
Yu, K., 80-160
Yuhr, L. B., 85-112; 86-465

Zamuda, C.. 85-412, 419; 86-457
Ziegenfus, L.  M., 84-521
Zicgler, F. G., 87-70; 85-349
Zimmerman, P. M., 84-326
Zuras, A. D., 85-1
556     AUTHOR INDEX

-------
                                             SUBJECT  INDEX
                     This  Subject  Index contains subjects presented in  1980-86 only.
Above Ground Closure, 83-215
Acidic Waste Site, 85-326
Activated Carbon, 57-374; 82-259, 262;
  55-209, 248, 253, 342
Adsorption,  54-393
Advanced Technologies, 54-412
Agency for Toxic Substances and Diseases
  Registry, 56-467
Agricultural  Fire Residue, 54-420
Air Modeling, 52-331; 54-66
Air Monitoring, 52-67, 268, 299, 306,  331;
  53-82, 85;  56-152
   Ambient, 57-280; 53-293; 55-125
   Cleanup Site, 54-72
   Design, 56-152
   Emissions, 52-70
   Nitrogen  Compounds, 53-100
   Real Time, 53-98
Sampling Techniques, 52-334; 56-152
   Two Stage Tube, 53-85; 54-81
Air Photos, 50-116; 55-116
Air Quality,  52-63
   Assessment,  52-76
Air Stripping, 53-209, 313, 354; 54-170
   Emissions Control, 54-176
   Soils, 56-322
Alternative Concentration Limits, 56-173
Alternative Treatment Technologies, 56-361
Analysis, 52-45
   Drum Samples, 54-39
   Metals, 53-79
   Mobile, 56-120
   Portable  Instruments, 52-36,  40, 57
   Pyrographic, 57-114
   Quality Control, 54-29
   Screening, 53-86; 55-97
   Site Data Base, 54-49
   Spectrometer, 53-291
Arizona
   TCE Contamination,  52-424
Arsenic Waste, 54-469; 55-109
Asbestos, 55-21
ASCE, 57-2
Assessment,  52-17, 27; 53-37
   Areal Photography, 55-116
   Biological, 52-52
   Cold Weather, 52-254
   Endangerment, 54-213, 226
   Environmental, 56-1
   Exposure, 56-69
   Health Effects, 54-253
   Health Risk, 54-230, 261
   Management, 57-348, 351
   Mathematical Modeling, 57-306, 313
   Mercury  Contamination, 52-81
   Methods, 57-79
   Pesticide Plant, 52-7
   Risk, 56-69
   Site, 55-209
Auditing, 57-398
Baird & McGuire Site, 55-261
Barriers, 52-249
   Bentonite, 52-191
   Cement, 54-126
   Gelatinous, 52-198
   Geomembrane, 56-282
   Leachate Compatibility, 54-131
   Sorptive Admix, 56-277
Basic Extraction Sludge Treatment, 56-318
Bedrock Aquifers, 55-142
   Contaminant Movement, 52-111; 55-202
Bench Scale Study, 57-288
Bench Scale Testing, 50-184
Beneficial Use, 54-560
Bentonite-Cement Mixtures
   Durability, 55-345
Bentonite-Soil Mixture Resistance, 54-131
Bentonite-Soil Slurry Walls, 55-357, 369
Berlin & Farro,  57-205
Bid Protests,  54-520
Bidding Cleanup Contracts, 54-509
Biodegradation, 52-203; 54-393; 55-234
Bioindicators, 57-185
Biological
   Monitoring,  57-238
   Treatment, 56-253
Bioreclamation, 55-239
Block Displacement Method, 52-249
Bottom Barrier, 54-135
Bridgeport Rental and Oil Services Site, 55-299
Bromine
   Organic, 52-442
Building Decontamination, 54-486
Buried Drums
   Sensing, 50-239
California
   Superfund Program, 52-428
   Ranking System, 55-429
Callahan Site, 52-254
Capping, 53-123, 296
   Cost, 53-370
Carcinogens,  54-11
   Reportable Quantities, 56-162
Cell  Model, 55-182
Cement/Asphalt Emulsion, 54-131
Cement/Bentonite Slurry Wall, 56-264
CERCLA (See Also Superfund)
   EPA/State Relations, 56-22
   Options and Liabilities, 56-18
   Remedies, 55-4
Change Orders, 54-521
Chemical
   Analysis
      Rapid, 50-165
   Control, 57-341; 54^16
   Oxidation, 53-253
   Plant
      Emergency Removal, 53-338
   Specific Parameters, 55-412
Chemometric Profiling, 56-242
Children
   Arsenic Exposure, 55-409
China, 54-604
Chlorinated Hydrocarbons
   Groundwater Monitoring, 52-1
Chromic Acid, 56-448
Chromium Sludge, 50-259
Circulating Bed Combustor, 55-378
Citizen Information Committees, 55-473
Claims, 54-521
Clay
   Leachate Interaction, 53-154
   Organic Leachate Effect, 57-223
Cleanup, 50-147, 257
   Air Monitoring, 54-72
   Asbestos, 55-21
   Assessment Role, 53-389; 55-116
   BT-KEMI Dumpsite, 53-342
   Case Studies, 53-395; 54-440
   Coal Tar, 53-331
   Cold Weather, 52-254
   Community Relations,  55-468
   Contract Bids, 54-509
   Cost Allocation, 54-326
   Cost-Effectiveness? 56-193
   Criteria, 53-301
   Delays, 53-320
   Drum Site,  53-354
   Dual Purpose, 53-352
   Enforcement, 54-478
   Extent, 53-433
   Federal, 55-7
   Federal/State Cooperation, 55-50
   Forced, 57-255
   Generator, 55-7
   Gilson  Site  Proposal, 52-289
   Hardin County Brickyard, 52-274
   Level, 53-398; 56-173
   Liability Due to Failure, 53-442
   Long-Term Effectiveness, 52-434
   Management, 53-370
   Pacific Island, 54-427
   PCB, 52-156, 284
   Picillo  Farm, 52-268
   Public  Information Needs, 54-368
   Radioactive Mine Tailings, 54-504
   Radium Processing Residues, 54-445
   Reserve Fund, 55-58
   Rocky  Mountain Arsenal, 55-36
   Staged Approach, 52-262
   State-of-the-Art Technology, 55-285
                                                                                                        SUBJECT INDEX     557

-------
    Toxic Wastes, 8,5-311
    Under Superfund, 86-407
 Cleve Reber Site. 85-136
 Closure, 87-259
    Copper Residue Disposal Site, 81-10
    Cost Apportionment,  86-56
    Creosote Impoundment, 55-323
    Impoundment, 83-195
    Industrial Site, 84-111
    In place. 64-185
    Vickery, OH, 86-297
 Closure/Post-Closure
    Illinois Perspective, 83-459
 CMA, 8/-1
 Coal Mine Groundwater Cleanup, 84-356
 Coal Tar Cleanup, 83-331; 84-11
 Community Coordinator,  87-411
 Community Relations (See Also Public
    Participation), 8/-405,  415; 82-354; 84-378
        Activities, 84-371
        Benefits,  85-31
        Health  Concerns, 82-321
        Program, 83-386, 389
 Community Right-to-Know,  85-11
 Compatibility Testing, 8M10
 Composting
    Soils, 82-209
 Comprehensive Environmental Assessment and
    Response Program, 86-1
 Computer Risk Analysis, 84-300
 Computerized Expert Systems, 86-208
 Connecticut
    Risk Evaluation, 80-25
 Consultant
    Liability, 86-47
 Containment System Design, 82-175
 Contaminant Transport,  86-88
 Contaminated Land, 84-549
 Contaminated Soil, 85-226, 231
    Cleanup. 83-354
 Contamination
    Mapping, 83-71; 84-85
 Contingency Fund, 80-21
 Contingency Plan
    Massachusetts, S3-420
    Remedial Sites, 84-489
 Contractors
    Indemnification, 86-52
 Contracts
    FIT, 86-36
    REM/FIT. 83-313
    Remedial Planning, 86-35
    Superfund, 86-40, 46
    Technical Enforcement Support, 86-35
 Cooperative Agreement, 84-103; 85-53
 Copper Smelter
    Arsenic Wastes, 85-409
 Cost, 80-202; 87-248; 82-289; 83-209
    Above Ground Waste  Storage, 82-228
    Air Stripping, 83-313
    CERCLA Financed, 85-395
    Cleanup, 82-262; 85-2%. 366, 370; 84-341
    Cleanup Allocation Model, 84-326
    Cleanup level. 83-398
    Closure Apportionment,  86-56
    Computer  Models, 83-362
    Cover, 82-187
    Discounting Techniques,  86-61
    Effective Screening, 85-93
    Effectiveness Evaluation, 82-372;
    84-290; 86-193
     Estimates, 80-202; 84-330, 335
     Ground Freezing,  84-386
     Groundwater Treatment, 83-248, 358
     Health and Safety Impact, 83-376
     Leachate Collection, 83-237
   Leachate Monitoring, 82-97
   Management, 84-339
   Minimization, 87-84
   Recovery, 84-313
   Recovery Documentation, 82-366
   Remedial. 82-118
   Savings. 86-164, 420
   Savings via Negotiation, 82-377
   Treatment System, 8/-294
   Water  Recovery System, 82-136
Coventry,  RI, 80-239
Covers (See Also Caps), 82-183, 187.
  448; 84-588
   Design and Construction, 85-331
   Pesticide Disposal Site, 85-349
Creep Characteristics, 86-247
Creosote Impoundment, 85-323
Criticism,  84-532
Cutoff Wall. 83-123, 296
   Chemically Rcsisiant, 83-169,  179, 191
   Cost, 83-362
Cyanides,  84-598, 600
Damage
   Cost Recovery, 8/-393
Data Bases. 83-304;  84^9, 59
   Problems, 86-213
Data Quality Objectives, RI/FS, 86-398
DC Resistivity, 86-227
Decision-Making, 8/-230
Decision Tree Analysis, 82-408
Decontaminating Buildings, 84-486
Decontamination, 80-226
   Waterway, 83-21
Degradation
   TNT Sludge, 83-270
   VOCs, 84-217
Denney Farm, 81-326
Department of Defense Program, 82-128
   TNT Cleanup, 85-314
Department of Energy, 85-29
Design
   Mathematical Modeling, 87-306
   Preliminary, 80-202
Detection
   Buried Drums, 84-158
Detonation, 84-200
Detoxification, 80-192; 84-382
   Fire Residues, 84-420
Discovery  Methods,  86-84
Diesel  Fuel, 86-415
DIMP. 8/-374
Dioxin. 8/-322, 326;  83-405;
  84-287;  85-261; 86-78, 97
Dispersion Coefficients, 83-135
Disposal,  87-329
   Above Ground, 83-275
   Commercial Criteria, 82-224
   Computer Cost Model, 83-362
   Liability, 83-431
   Mine,  85-387
   Salt Cavities, 83-266
   Shock  Sensitive Chemicals. 84-200
Documentation
   Cost Recovery, 82-366
DOD
   Hazardous Materials Technical Center,
  82-363
   IRP, 85-26
   Site Cleanup, 83-326
DOE
   CEARP, 86-1
Downhole Sensing, S3-108
Drain  System, 83-237
Drainage  Nets, 86-247
Drilling
   Buried Drum Pit, 86-126
   Horizontal, 86-258
Drinking Water Contamination, 84-600
Drums, 82-254
   Analysis, 84-39
   Buried, 82-12; 84-158
   Disposal  Pit,  86-126
   Handling, 82-169
   Site Cleanup, 83-354
Dust Control, 84-265
Electric Reactor. 84-382
Electromagnetic
   Induction, 83-28, 68; 86-132, 227
   Resistivity, 82-1
   Survey, 80-59; 82-12
   Waves, 80-119
Emergency
   Planning, 84-248
   Removal, 83-338
Emissions
   Monitoring, 83-293
   Rates, 84-68
Endangerment Assessments, 84-213;
  85-3%. 423. 438
Enforcement, 84-544; 85-21
   CERCLA
       EPA/State Relations. 86-18
   Cleanup. 84-478
   Endangerment Assessments. 84-213; 85-3%
   Information  Management.  85-11
Environmental
   Concerns, 84-592
   Impact, 87-177
   Risk Anah sis. 82-380
Epidemiologic Study, 84-287
   Dioxin, 86-78
Excavation,  82-331
Exhumation. 82-150
Expedited Response Action Program, 86-393
Exploratory  Drilling, 86-126
Explosives
   Contaminated Soils  Incineration, 84-203
   Waste Disposal Sites, 84-141
Exposure
   Assessment, 86-69
   Children, 84-239
   Response Analysis, 82-386
Extraction, 84-576
Fast-Tracked Hydrogeologkal Study, 85-136
Feasibility Study
   Arsenic Waste,  84-469
Federal Cleanup, 85-7
Federal Facility  Coordinator, 85-32
Federal/State Cooperation, 82-420; 85-50
Field
   Data Acquisition, 86-148
   Identification, 85-88; 86-120
   Quality Assurance, 86-143
   Sampling, 84-85, 94
   Screening, 86-105
Fire, 87-341; 82-299
   Underground, 86-350
Firefighter
   Toxic  Exposure, 86-152
First Responder Training,  85-71
FIT
   Contracts, 83-313; 86-36
   Health and Safety, 80-85
Fixation/Solidification, 86-297
Floating Covers, 84-406
Florida s  Remedial Activities, 82-295
Fluorescence, 86-370
Fort Miller,  87-215
Foundry Wastewater, 84-598
Fractured Bedrock. 84-150
FT/IR. 86-371
Fugitive
558     SUBJECT INDEX

-------
   Dust Control, 54-265
   Hydrocarbon Emission Monitoring, 81-123
Gas Chromatograph, 82-51, 58; 83-76
   Portable, 82-36; 83-105
   Screening, 86-386
Gas Plants, 86-93
GC/MS, 82-57
Gases
   Collection and Treatment, 86-380
   Unknown, 54-416
Gasification Plant Site  Contamination, 86-242
Gasoline, 55-269
Generator Cleanup, 85-7
Geographic Information Systems, 55-200
Geohydrology, 55-117
Geomembranes, 86-269
   Barrier Technology, 55-282
   Seam Testing, 55-272
Geophysical, 55-68,  71
   Investigation, 54-481; 55-217
   Logging, 55-292
   Methods, 52-17
   Modeling, 55-110
   Monitoring, 55-28
   Survey, 57-300
   Techniques, 55-130; 55-465
Geophysics, 57-84; 52-91
   Characterizing Underground Wastes, 55-227
Geostatistical Methods, 55-107; 55-217
Geotechnology
   Containment  System, 52-175
   Property Testing, 55-249
   Techniques, 55-130
Germany, 54-565, 600
Gilson Road Site, 52-291
Ground Freezing, 54-386
Ground Penetrating Radar, 50-59, 116, 239;
57-158, 300; 55-68; 55-227 Groundwater
   Activated Carbon Treatment, 55-361
   Alternatives to Pumping, 52-146
   Applied Modeling, 55-430
   Bedrock Aquifers, 55-403
   Biological Treatment, 55-253, 333
   Biodegradation, 55-234
   Case Histories, 55-430
   Chrome Pollution, 55-448
   Cleanup, 52-118, 159; 55-354; 54-176
   Collection, 55-220
   Cyanide Contamination, 54-600
   Containment, 52-259; 55-169
   Containment Movement, 52-111; 55-147
   Contamination, 57-329,  359; 52-280;
   55-43, 358; 54-103, 141, 145,  162
  170, 336; 55-43, 157, 261
      Detection, 54-20
      Liabilities, 55-437
      Mapping,  55-71
      Potential,  50-45
   Flow System, 55-114, 117
   Flushing, 55-220
   Halocarbon Removal, 55-456
   Heavy Metals, 55-306
   HELP, 55-365
   Horizontal Drilling, 55-258
   Hydraulic Evaluation, 55-123
   Hydrologic Evaluation of Landfill
   Performance, 55-365
   Investigation, 50-78, 54-1, 107; 55-158
   In  Situ Biodegradation, 55-239
   Lime Treatment, 55-306
   Mathematical Modeling, 57-306
   Metal Finishing Contamination, 55-346
   Microbial Treatment, 55-242
   Migration, 50-71; 54-150, 210
      Prevention, 55-179,  191; 54-114; 55-277
   Mobility, 54-210
   Modeling, 52-118; 55-135, 140, 145;
   54-145; 55-88
   Monitoring, 50-53; 52-17, 165
       Evaluation, 55-84
       Interpretation, 52-86
       Long-Term, 55-112
       Statistics, 54-346; 55-130
       Well Design and  Installation, 55-460
   Plume Definition, 55-128
   Pollution Source, 57-317
   Post-Closure Monitoring, 55-446
   Protection, 50-131, 54-565
   Recharge, 55-220
   Recovery Cost, 52-136
   Recovery Design, 52-136
   Remedial Plans, 55-130
   Remediation, 55-220
   Research Needs, 55-449
   Restoration, 52-94; 54-162; 55-148
   Sampling, 57-143, 149
   Slurry Wall, 55-264
   Studies, 55-431
   Superfund Protection Goals, 55-224
   TCE Contamination, 52-424
   Treatability, 57-288
   Treatment, 50-184; 52-259;
   55-248, 253; 55-220
   Ultra Clean Wells, 55-158
   VOC Biodegradation, 54-217
Grout, 55-169, 175
   Chemistry, 52-220
Grouting, 52-451
   Silicates, 52-237
Halocarbon Removal, 85-456
Halogen
   Combustion Thermodynamics, 55-460
Harrisburg International Airport, 55-50
Hazard
   Degree, 57-1
   Potential, 50-30
   Ranking, 57-188
       Prioritizing, 57-52
       Scoring, 55-74
       System, 57-14; 52-396
       U.S. Navy Sites, 55-326
   Unknown, 57-371
   vs Risk, 54-221
Hazardous Materials
   Identification, 55-88
   Storage
       Spills, 52-357
   Technical Center, 52-363
Hazardous Waste
   Emergencies
       Information Sources, 54-59
       In situ Vitrification, 55-325
   Expert Management System, 55-463
   Land Treatment,  55-313
   Management  Facility Siting, 54-517
   Policies, 54-546
   Screening, 55-370
   Site Reuse, 54-363
   Treatment, 55-303
Health and Safety (See Also Safety)
   Assessments,  54-261, 55-423
   Community Concerns,  52-321
   Cost Impact, 55-376
   Guidelines, 55-322
   Hazards, 50-233
   Plan, 55-285
   Program, 50-85, 91,  107
   Training, 55-473
Health Risk Assessment, 54-230,  253
Heart Stress Monitoring, 54-273
Heavy Metals
   Impoundment Closure, 55-195
    X-Ray Fluorescence, 55-114
 Herbicide Mixing, 55-97
 High-Pressure Liquid Chromatography, 55-86
 Horizontal Drilling, 55-258
 Hyde Park, 55-307
 Hydrocarbons, 55-269
    Biodegradation, 55-333
    Leaks, 52-107
    Recovery, 55-339
 Hydrogeologic
    Data, 54-6
    Evaluation, 50-49
    Fast-Track, 55-136
    Landfill, 55-182
    Investigation, 57-45, 359; 55-346;
    55-148, 403
 Identification, 55-63
    Hazardous Material, 55-88
    Reactivity, 55-54
 Illinois
    Closure/Post Closure,  55-459
 Immobilization, 52-220
 Impact Assessment, 57-70
 Impoundment, 50-45
    Closure, 55-195; 54-185;
     55-323; 55-318
    Leaks, 55-147
    Membrane Retrofit, 52-244
    Sampling, 55-80
 Incineration, 52-214; 55-378, 383
    Explosives  Contaminated Soils, 54-203
    Halogens, 55-460
    Mobile, 50-208; 57-285
    Performance Assessments, 55-464
    Research, 54-207
    Safety, 55-4
    Sea, 50-224
 Indemnification, 55-52
 Inductive Coupled Plasma Spectrometer, 55-79
 Information
    Committees, 55-473
    Management, 55-11
 Infrared  Incinerator, 55-383
 In Situ
    Biodegradation, 55-234, 239, 291
    Chemical Treatment, 55-253
    Pesticide Treatment, 55-243
    Solidification/Fixation, 55-231
    Stabilization, 55-152
    Treatment, 54-398; 55-221
    Vitrification, 54-195
 Installation Restoration Program
    McClellan AFB, 54-511; 55-26
 Insurance, 52-464
 Interagency Management Plans, 50-42
 Investigation
    Hydrogeologic, 52-280
Kriging, 50-66
Laboratory
    Management, 57-96
    Mobile, 55-120
    Regulated Access, 57-103
La  Bounty Site, 52-118
Lagoons,  57-129; 52-262
   Floating Cover, 54-406
Landfill
    Closure, 50-255
   Covers, 55-365
    Future Problems, 50-220
    Risk,  55-393
Land Treatment, 55-313
Leachate
    Clay Interaction, 55-154
   Characterization, 55-237
   Collection,  55-237; 55-192
    Control, 54-114; 55-292

                  SUBJECT INDEX     559

-------
     Drainage Nets, 56-247
     Effects on Clay,  57-223
     Generation Minimization, 80-135, 141
     Migration, 52-437; 84-211
     Minimization, 81-201
     Modeling, 55-135; 84-91;
       55-189
     Monitoring Cost, 52-97
     Plume Management, 55-164
     Synthetic, 5(5-237
     Treatment, 50-141; 52-203, 437; 55-202,
       217; 54-393; 55-192
  Lead. 54-239; 55-442; 56-164. 200, 303
  Leak Detection,  55-94, 147; 55-362
  Legal Aspects
     Extent of Cleanup, 55-433
  Legislation
     Model Siting Law, 50-1
  Liability, 52-458, 461, 464, 474
     Consultant, 56-47
     Corporate, 50-262
     Disposal, 55-431
     Generator, 57-387
     Groundwater Contamination, 55-437
     Inactive Sites, 50-269
     Superfund Cleanup Failure, 55-442
     Superfund Minimization, 56-18
     Trust Fund, 55-453
  Liner
     Breakthrough, 55-161
     Flexible, 54-122
     Leak Detection,  55-362
     Leak Location, 52-31
     Synthetic Membrane, 55-185
     Testing, 56-237
  Love Canal. 50-212, 220; 57-415; 52-159,
    399; 56-424
  Low Occurrence Compounds, 55-130
  Magneuometry,  50-59, 116; 5/-300;  52-12;
    55-68; 56-227
  Management Plans
     New Jersey, 55-413
  Managing Conflict,  54-374
  Mass Selective Detector, 55-102
  Massachusetts Contingency Plan. 55-420; 55-67
  McClellan AFB. 55-43
  Medical Surveillance, 54-251, 259; 56-455
  Mercury, 52-81
  Metals, 52-183
     Analysis, 55-79
     Detection, 50-239
     Detector, 50-59;  57-300; 52-12
     Finishing, 55-346
     Screening, 55-93
  Microbial Degradation, 55-217, 231. 242
  Microdispersion, 54-398; 55-291
  Migration, 54-588
     Cutoff,  52-191
     Prevention, 52-448
  Mine
     Disposal, 55-387
     Mine/Mill Tailings, 55-107
     Sites, 55-13
     Tailings Cleanup, 54-504
     Waste Neutralization and Attenuation,
       56-277
  Mobile
     Incinerator, 55-378, 382
     Laboratory, 50-165; 54-45; 56-120
     MS/MS, 54-53
     Treatment, 56-345
  Modeling
     Applied, 56-430
     Cell, 55-182
     Geophysical Data, 56-110
     Groundwater Treatment, 55-248
   Leachate Migration, 52-437; 55-189
   Management Options, 55-362
   Remedial Action, 55-135
   Site Assessment. 57-306
Monitoring
   Ambient Air, 57-122,  136
   Wells
      Installation, 57-89
      Integrity Testing, 56-233
      Installation In In-Place Wastes, 56-424
      Location, 57-63
MS/MS Mobile System, 54-53
Multi-Site/Multi-Activity  Agreements, 55-53
National
   Contingency Plan Revisions, 56-27
   Contract Laboratory Program, 54-29
   Priority List (NPL), 55-1
      Deletion, 56-8
      Mining Sites. 55-13
   Resource Damage, 5/-393
   Response. 5/-5
NATO/CCMS Study, 54-549
Natural Resources Restoration/Reclamation,
  54-350
Negotiated Remedial Program, 54-525
Negotiating, 52-377, 470
Netherlands. 54-569
Neutral Validation RI/FS, 56-445
Neutralization, 55-63
New Jersey
   Cleanup Plans, 55-413
   DEP, 55-48
   Reserve Fund, 55-58
New York City, 54-546
NIKE Missile Site Investigation, 56-436
No-Action Alternative, 55-449
Non-Destructive Testing Methods, 52-12,
54-158;  56-272
North Hollywood Site, 54-452
Occupational Health Programs, 54-251, 259
Odor, 52-326; 55-98
Oil
   Pond Pollution,  56-415
   Recovery, 55-374
   Sludge
      Best, 56-318
Old Hardin County Brickyard, 52-274
Olmsted AFB, 55-50
OMC Site, 54-449
On-Site Leachate Renovation, 54-393
Organics
   Emissions, 52-70, 54-176
   Land Treatment, 56-313
   Sludge Stabilization. 54-189
   Solvents Permeability, 54-131
   Vapor
      Analysis, 55-98
      Field Screening, 55-76
      Leak Detection, 55-94
      Personnel Protection, 57-277
   Wastes
      Characterization, 54-35
Ott/Story, 57-288
Pacific Island Removal. 54-427
Parametric Analysis, 57-313
PCBs, 57-215; 52-156, 284; 55-21, 326,
  366. 370; 54-243,  277, 449; 56-420
   Field Measurement, 55-105
   Screening, 56-370
Pennsylvania Program, 57-42
Permeability Coefficient Measurement, 54-584
Personnel
   Protection Levels,  57-277
   Safety Equipment, 56-471
Pesticides, 52-7; 55-255, 349; 56-386
   In Situ Treatment, 55-243
   Risk Assessment, 56-186
Petro Processors Site, 54-478
Petroleum Contamination, 54-600
Photographic Interpretive Center, 54-6
Physical Chemical Data Use, 54-210
Picillo Farm Site, 52-268
Pilot Plant, 57-374
PIRS,  52-357
Pittson, PA. 50-250
Plan Review, 56-143
Plant Bioindicators, 57-185
Pollution Abatement Site, 54-435
Polyaromatic Hydrocarbons, 54-11
Polynuclear Aromatic Hydrocarbons, 56-242
Post-Closure
   Care, 57-259
   Failure, 55-453
   Groundwater Monitoring, 55-446
   Monitoring, 52-187
   Monitoring Research, 55-449
Potentially Responsible Party (PRP), 55-275
POTW
   Leachate Treatment, 55-202
Price Landfill
   Remedial Action, 55-358
Priorilization (See Also Hazard Ranking),
  57-188
Private Cleanups at Superfund Sites, 56-27
Private Property Legal Issues,  56-31
Probabilistic Spatial Contouring, 55-442
Public
   Awareness, 55-383
   Health, 54-232; 55-438
   Information Program, 50-282; 54-3; 55-473
       Needs, 54-368
   Involvement, 55-476
   Participation (See Also Community
     Relations>52-340. 346,  350; 55-383
       Failures, ฃ5-392
   Policy
       Cleanup  Level, 55-398
   Relations, 55-468
Pulsed Radio Frequency, 57-165
Quality
   Assurance
       Audits, 54-94; 56-143
       Monitoring  Well Integrity, 56-233
   Control, 52-45; 54-29; 56-287
Radar  Mapping, 55-269
Radioactive
   Mine Tailings. 54-504
   Site Assessment, 55-432
   Wastes. 57-206
Radionuclides, 56-306
Radium Processing Residues, 54-445
Radon
   Contamination, 54-457
   Gas, 52-198
RAMP, 52-124
   Love Canal, 52-159
Ranking System. 57-14;  55-429 RCRA
   Requirements, 55-4
   Section 3012, 54-535, 544
   Superfund Interrelationship, 56-462
RDX,  52-209
Reactivity
   Identification, 55-54
Real Estate
   Hazardous Waste Implications, 52-474
Reclamation
   Chromium Sludge, 50-259
Records Management System,  57-30
Recovery
   Hydrocarbons, 56-339
   Organics, 54-145
Regional Response Team, 50-6; 52-274
560     SUBJECT INDEX

-------
REM Contracts, 83-313
Remedial
   Action, 52-289
      Alternatives, 84-35, 277, 290, 306,
        321; 86-361
           Risk Assessment, 55-319; 86-65
      Bedrock Aquifers, 86-403
      Case Studies, 82-131
      Contingency Plans, 84-4S9
      Costs, 84-335, 341
      Cost Management, 54-339
      Decision-Making, 84-66
      Design, 50-202
      Pesticides,  55-255
      Florida's Site, 52-295
      Groundwater, 54-565
      Investigation, 54-435
      Guidance, 54-498
      Lessons, 54-465
      Negotiated, 54-525
      Netherlands, 54-569
      North Hollywood Site, 54-452
      Options, 50-131
      Pesticides,  55-186
      Planning, 55-281
      Planning Contracts, 56-35
      Priority System, 55-432
      Progress Status, 50-125
      Public Involvement,  55-476
      Screening and Evaluation, 54-62
      Selection, 54-493
      Smelter Site, 86-200
      Technologies, 55-285
   Construction
      Safety Plans, 55-280
   Cost Estimation Model, 54-330
   Design
      Groundwater, 53-123, 54-109, 356
      Model Based Methodology, 55-135
      OMC Site, 54-449
      Thamesmead, 54-560
   Projects
      Corps of Engineers,  53-17
   Response
      Role of U.S. Army,  52-414
   Technologies
      Screening and Evaluation, 54-62
Remediation
   Discounting Techniques, 56-61
   Innovative Approach, 55-307
Remote Sensing, 50-59, 239; 57-84, 158,
  165, 171
Reportable Quantities, 56-182
Research
   Post-Closure Monitoring,  53-449
   U.S. EPA Program, 50-173
Reserve Fund, 55-58
Resistivity, 50-239;  57-158;  52-31; 53-28
Resource Recovery, 57-380
Response
   Model, 57-198
   Procedures, 50-111
Restoration
   Swansea Valley, 54-553
Reusing Hazardous Waste Sites, 53-363
Reverse Osmosis, 52-203
RI/FS
   Bridgeport Oil and Rental Services Site,
     55-299
   Chromic Acid Leak, 56-448
   Computerized Expert Systems, 56-208
   Data Quality Objectives, 56-398
   Neutral Validation, 56-445
   NIKE Missile  Site, 56-436
   Wood  Treating Site, 56-441
Right-to-Know, 56-4
Risk
   Acceptability, 53-405
   Analysis, 57-230; 53-37
      Computer, 54-300
      Environmental, 52-380
   Assessment, 57-238; 52-23, 386, 390, 406,
  408; 53-342; 54 283, 321; 55-393, 412,
  449; 56-69, 74, 457
      Air Quality, 52-63
      Comparative,  53-401
      Data Problem, 56-213
      Health, 54-230
      Manual, 55-419
      Modeling, 52-396
      Prioritizing, 55-433
      Quantitative, 54-290; 56-65, 186
      Remedial Action Alternatives,  55-319
      Underground Tanks, 54-16
      U. S. EPA Guidelines, 56-167
   Cleanup Level, 53-398
   Decision Analysis Module, 56-463
   Design, 54-313
   Evaluation, 50-25
   Minimization, 57-84
   Perception, 56-74
Rocky Mountain Arsenal, 57-374; 52-259;
  55-36
Safety (See Also Health and Safety),  52-299,
  306; 55-406
   Cost Impact, 52-311
   Equipment, 56-471
   Incineration, 56-4
   Information, 54-59
   Plans, 54-269
   Procedures, 57-269
   Remedial Construction, 53-280
   Sampling and Analysis, 57-263
   Tank Investigation and Removal,  55-198
   Training, 52-319
Sample Thief,  57-154
Sampling, 50-91
   Analysis
      Safety,  57-263
   Biological,  52-52
   Drums, 57-154
   Impoundments, 55-80
   Screening,  57-103, 107, 114
   Statistical-Based,  56-420
   Strategy, 55-74
   Subsampling, 54-90
   Techniques, 57-143, 149
Screening
   Analytical, 55-97
   Field, 56-105
   Mass Selective Detector, 55-102
   Metals, 55-93
   PCB, 56-420
   Spectrometry, 53-291
   Statistical,  56-164
   X-Ray Fluorescence, 56-115
Security, 53-310
Seismic
   Boundary Waves, 55-362
   Refraction, 50-239; 56-227
Sensing
   Downhole, 53-108
Serum Reference Methods, 54-243
Settlement, 55-275
   Agreements, 52-470
      Hyde Park, 55-307
Shenango, 50-233
Shock Sensitive/Explosive Chemical
  Detonation, 54-200
Shope's Landfill Cleanup, 53-296
Silicates, 52-237; 56-303
   Grouts, 53-175
Silresim Site, 52-280
Site
    Assessment, 50-59, 91; 53-221; 54-221;
     55-209
    Discovery, 53-37; 56-84
    Evaluation, 50-25, 30
    Hazard Rating, 50-30
    Inspection Sampling Strategy, 55-74
    Investigation, 55-48
    Location, 50-116; 57-52
    Location Methodology, 50-275
    Problems
      Whales, 54-594
    Reuse, 54-363, 560
Site Program, 56-356
Siting, 50-1
    Hazardous Waste Management Facility,
     54-517
    Public Information Needs, 54-368
Sludge
    B.E.S.T. Process, 56-318
    Stabilization, 56-277
Slurry
    Trench, 52-191
    Wall, 55-357, 374; 56-264
Small Quantity Generator, 55-14
Smelter
    Lead, 54-239; 55-442
    Site Remediation, 56-200
Soil
    Advanced Technologies, 54-412
    Air Stripping, 56-322
    Chemistry of Hazardous Materials, 56-453
    Contamination, 52-399, 442; 53-43;
     54-569,  576
      International Study, 52-431
      Pesticides, 55-243
    Cover, 56-365
    Extraction, 52442
    Gas
   Analysis, 56-138
    Sampling, 54-20
    Geotechnical Property Testing, 55-249
   Thermal Treatment, 54-404
    Vapor Measurement, 55-128
   Washing, 55-452
Soil-Bentonite Slurry Walls, 55-357,  369
Solid Waste Management
    China, 54-604
Solidification, 57-206
   Fixation, 56-247
   Organics, 56-361
    Silicates, 52-237
   TNT Sludge, 53-270
Solvent Mining, 53-231
Spatial Contouring,  55442
Spills
    Hazardous Materials Storage, 52-357
Stabilization, 50-192
    Viscoelastic Polymer Waste, 55-152
Stabilization/Solidification, 50-180;
  55-214, 231
    Organic Sludge, 54-189
    Quality Control, 56-287
State
    Criticism, 54-532
    Enforcement, 54-544
    Participation, 52-418; 54-53
    Plans
      New Jersey, 53-413
      Pennsylvania, 57-42
    Superfund Program, 52-428; 55-67
Statistical Methods, 54-243
    Groundwater Monitoring, 54-346; 56-132
    Sampling, 56-426
    Screening,  56-64
                                                                                                               SUBJECT INDEX     561

-------
Statistical Modeling
   Geophysical Data, 86-110
Steam Stripping, 82-289
Stringfellow Site, 80-15, 21
Subsampling, 84-90
Subsurface Geophysical Investigation, 54-481
Superfund (See Also CERCLA)
   California, 81-31
   Cleanup Failure Liability, 83-442
   Contractor Indemnification, 86-56
   Contracts, 86-40, 46
   Drinking Water, 83-8
   Federal/State Cooperation,  87-21; 83-426
   Groundwater Protection Goals, 86-224
   Impact on Remedial Action, 86-407
   Implementation, 83-1
   Innovative Technology Programs, 86-356
   Management, 83-5
   Private Cleanup, 86-27
   Private Property Cleanup, 86-31
   Private Sector Concerns, 87-10
   Programs
      New Jersey, 83-413
      Texas, 83^23
   RCRA Interrelationship, 86-462
   Revisited, 86-412
   Right-to-Know, 86-11
   Site  Management, 86-14
   State Perspective, 84-532
   Strategy for Dealing With, 86-469
   U. S. EPA Research, 8/-7
Surface
   Sealing, 8/-201
   Water Management, 80-152
Swansea Valley, 84-553
Swedish Dump Site Cleanup, 83-342
Sweeney, 82-461
Sydney  Mine Site, 85-285
Sylvester Site, 87-359
Synthetic Membrane Impoundment Retrofit,
  82-244
Tailings, 85-107
Tank Investigation and Removal, 85-198
TAT
   Health and Safety, 80-85
Technical Enforcement Support Contract,
  86-38
Technology Evaluation, 82-233
Texas
   Ambient Air Sampling, 85-125
   Superfund Program, 83-423
Thamesmead, 84-560
Thermal Treatment
   Soils, 84-404
Thermodynamics
   Halogen Combustion, 85-400
Thin-Layer Chromatography,  86-420
TNT, 82-209; 85-314
Top-Sealing, 80-135
Town Gas, 84-11; 86-93
Toxic Substances and  Disease  Registry Agency,
  85-403
Toxicological Data, 86-193
Trace Atmospheric Gas Analyzer, 83-98, 100
Training
   First Responders, 85-71
   Resources,  83-304
Treatment
   In Situ, 82-451; 83-217. 221. 226, 231
   Mobile, 86-345
   On-Site. 82-442
   System  Design,  81-294
Underground Tank
   Fuel, 86-350
   Spill Risk Assessment, 84-16; 86-176
   Trichloroethylene,  86-138,  430
   Waste Characterization, 86-227
United Kingdom, 80-8, 226
Unknown Gases, 84-416
U. S. Army Corps of Engineers, 82-414; 83-17
U.S. Army Installation Restoration Program,
  84-511
U. S. Coast Guard, 80-6
U. S. EPA
   Expedited Response Action Program,
     86-393
   Mobile Incinerator, 87-285
   Reportable Quantities, 86-182
   Research, 8/-7
   Risk Assessment Guidelines,  86-167
U. S. Navy, 85-48
UV/Ozone Study 85-456
Vados Zone Monitoring, 82-100
Vapor
   Emission, 82-326
   Soils, 85-128, 157
VUcoetastic Polymer Waste, 85-152
Vitrification
   In Situ, 84-191; 86-325
Volatile
   Nitrogen Compounds  Monitoring, 83-100
   Organic*
      Emissions, 8/-129; 84-68, 77
      Monitoring, 8/-I22; 84-12
      Screening, 86-386
      Stripping From Soils. 86-322
Wales. 84-594
Walls
   Design and Installation, 86-460
   Gelatinous, 82-198
   Slurry, 82-191
Waste Storage
   Above Ground, 82-228
Wastewater Treatment, 80-160; 84-598
Water Treatment
   Cost, 83-370
Waterway Decontamination, 83-21
West Germany, 83-68
WET Procedure, 86-303
Wetland Contamination, 85-261
Wilsonville Exhumation, 82-156
Wobum. MA, 87-63, 177
Wood Treating
   Facility, 87-212
   PAH. 86-242
   RI/FS, 86-441
X-Ray
   Analyzer. 85-107
   Fluorescence. 85-93; 86-115
Zinc, 86-200
562     SUBJECT INDEX

-------